Post on 12-Aug-2020
IOSR Journal of Business and Management (IOSR-JBM)
e-ISSN: 2278-487X, p-ISSN: 2319-7668. Volume 22, Issue 4. Ser. VIII (April. 2020), PP 31-71
www.iosrjournals.org
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 31 | Page
Effect of Corporate Governance, Investment Strategyand
Macroeconomic Factors on Financial Performance of Pension
Schemes in Kenya
AkwimbiAmbaka William A Thesis Proposal Presented In Partial Fulfillment Of The Requirements For The Award Of The
Degree Of Doctor Of Philosophy In Business Administration,
School Of Business, University Of Nairobi
Abstract: Pension schemes form a significant part of the global investment portfolio. In Kenya, they hold over
13% of the country’s GDP (OECD, 2018). Their importance is underscored by the fact that they contribute
significantlyto growth and development of world economies (Kakwani, Davis, 2005; Heijdra, Ligthart & Jency,
2006). Their financial performance is critical to the provision of retirement benefits. Khan, Nouman & Imran
(2015) observed that the financial performance indicates measures to which economic goals of an organisation
has been accomplished over particular time period. Pension schemes however, face numerous challenges that
can render the generation of retirement benefits inadequate.
A number of studies have been undertaken to evaluate the impact of factors that influence performance of
pension funds resulting in mixed and sometimes inconclusive findings. This study sought to assess the effect of
corporate governance, investment strategy, interest rate, inflation rate, exchange rate and GDP growth rate on
performance of pension funds in Kenya.The study was done using annual data on pension funds and economic
indicators spanning the period 1997 to 2018. In addition, it used questionnaires to gather data on corporate
governance and investment strategy indices.
Quantitative and correlational research design using Linear regression model was used to assess the effect of
corporate governance, investment strategy, interest rate, inflation rates, GDP growth rates and exchange rate
on pension performance. The study findings show that these factors had significant impact on pension funding.
They however, varied on their individual contribution to the prediction of funding level of each pension fund.
The study concludes that pension fund management and policy makers should take into consideration the effects
of macroeconomic factors, corporate governance and investment strategy in decision making on investment
plans to ensure generation of adequate funds to fulfill their key objective of providing retirement benefits to the
members.
----------------------------------------------------------------------------------------------------------------------------- ----------
Date of Submission: 25-04-2020 Date of Acceptance: 08-05-2020
------------------------------------------------------------------------------------------------------------------ ---------------------
I. Introduction
1.1 Background of Study Pension schemes form a significant part of the global investment portfolio. Their importance is
underscored by the fact that they contribute significantlyto growth and development of world economies
through provision of financial security post retirement, growth of investment, banking and insurance services as
well as development of capital markets (Kakwani, Davis, 2005; Heijdra, Ligthart&Jency, 2006; Watson, 2007;
Yermo, 2008). In Kenya, pension funds hold over 13% of the country’s GDP (OECD, 2018).
The financial performance of some pension funds worldwide has however deteriorated following
regional market crisis and large corporate failures such as the Asian financial crisis of late 1997 (Nam & Nam,
2004), the Global financial crisis of 2008 (Antolín & Stewart, 2009),and the tragic collapses and losses of giant
companies including the Enron Corporation (2001) in the US, the Bank of Credit and Commerce International
(BCCI) (1991) in the UK (Kaur &Suveera, 2009)among others.These events led to major reductions in pension
fund assets worldwide (OECD 2008), exacerbating the threat of pension funds failing to provide retirement
benefits (Besley & Prat, 2005). In Kenya, similar challenges were witnessed in the past two decades
particularly, operational malpractices, misappropriation of scheme funds and lack of transparency. The situation
was aggravated by the poor performance of the economy. The relationship between CG and firm value is widely
documented particularly in developed countries, but there is a clear gap in academic research in developing
countries on this issue.
Governance is progressively acknowledged as critical to the proper functioning of pension plans. A
growing body of literature in finance provides evidence on the association between CG practices, corruption,
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 32 | Page
legal framework and firm performance and value (Ficici&Aybar, 2012; Clark &Urwin, 2008; Moriarty
&Zadorozny, 2008). Carmichael and Palacios (2003) argues that good governance practices help to mitigate
conflicts among stakeholders in pension funds. Various studies have used several proxies to evaluate the effect
of CG rules and practice on business success or failure. Such proxies include board size, board independence,
audit committee, CEO duality and scrutiny of managers and executives by various stakeholders which
traditionally are not responsible or bothered to take this sort of governance duties (Kacperczyk, 2009). Chow
(2005), Yang and Mitchell (2008), and Manuel and Andreas (2008) found positive links between good CG and
performance. The findings are based on the theory that an efficient board of directors can significantly reduce
agency costs. Other scholars such as Larcker, et al. (2007), Bhagat and Black (2002) and Heracleous (2001)
however, found mixed and sometimes inconclusive results on the relations between CG and firm performance.
Evidence of the financial effect of CG on pension performance in developing countries is still relatively scarce.
Locally, studies focusing on the relationship between CG and stock performance of NSE listed firms established
that there is a link between the two factors (Ongore & Kobonyo 2011; Mirig’u 2011; Lishenga 2012). None
however, focused on pension funds. Thus, the mixed and sometimes inconclusive empirical literature from
developed countries in addition to the limited studies on the subject in developing countries creates a need for
further research.
Investment strategy is another key factor that is postulated to affectpension performance. Tonks (2006)
affirms that the investmentdecisions promote both the performance and the financial security of pension plan
benefits. It is therefore critical that this function is implemented and managed responsibly. This however, was
only recentlyrealised following financial crises and large corporate failures that led to decline in pension assets
worldwide (Stewart, 2010). Markowitz’s (1952) theory of investment in financial assets, the Modern Portfolio
Theory (MPT), is key in managing these risks. The theory is concerned with risk and return, based on the mean-
variance efficiency for assets allocation. It states that diversification of idiosyncratic risk has a relationship with
the expected rates of return on securities. Itthus provides a mechanism to find the optimal combination of
securities having minimum variance, the efficient portfolio. The investment environment in the developed world
varies from that in the developing countries. A study to evaluate the effect of investment strategy and CG on
financial performance of pension funds will therefore extend and strengthen knowledge on the subject from a
developing countries perspective.
Macroeconomic factors effect on financial performance of pension assets is a major factor of
consideration for long-term investors (Brinson et al., 1991). Ross’s (1976) equilibrium Asset Pricing Theory
states that the expected security returns is a function of multiple factors. Chen et al. (1986), Flanery and
Protopapadakis (2002) and Singh (2010) identified changes in Gross Domestic Product (GDP), inflation, risk
premium and exchange rate as well as legal and regulatory environment as the factors that influenced stock
market returns. Being major investors in financial assets, pension funds are likely to be affected by these factors,
as their financial performance will be determined by portfolio performance. Several study findings affirm that
macroeconomic factors influenced stock market returns in the developed world and Emerging Market
Economies (Mookerjee & Yu, 1997; Fama & French, 1989; Kwon & Shin, 1999). Other studies, however find
that over time the influence of certain macroeconomic factors on financial performance of stock is uneven
(Mohammed & Rasheed, 2002; Singh, 2010). Limited studies are available in the country on their effect on
pension performance. There are however, local studies involving other sectors showing that interest rates,
inflation rates, real GDP influenced stock returns (Olweny & Omondi, 2011; Ochieng & Oriwo, 2012; Osoro,
2015). None was on pension funds.
The study is anchored on the Agency theory of Jensen and Meckling (1976). It will however, be
supported three other theories namely the Modern Portfolio Theory (MPT) the Arbitrage Pricing Theory (APT)
as well as the Stakeholder Theory (SHT)).The Agency concept postulates that there is a relationship between
organizational structure and firms’ financial performance. The theory seeks to resolve conflict of interest and
agency costs that arise as a result of variation in risk preferences, information failure and shareholders having
minimal influence in decision-making in the firm, a role left to the management. Marashdeh (2014) postulated
that reduced agency problems raise share value leading to improved performance. The four theories are
integrated to develop testable hypotheses that will investigate the relationship between CG, investment strategy,
macroeconomic factors and pension financial performance in Kenya.
1.1.1. Corporate Governance
Corporate governance (CG) is described as systems and processes by which an organization
accomplishes its undertakings with the goal of mitigating conflicts among its stakeholders and get the best out of
their wellbeing (Carmichael & Palacios, 2003). The International Organization of Pension Supervisors (IOPS)
refers to pension governance as the framework by which the management makes choices about the pension
fund’s undertakings. It comprises the configuration of the board; the decision-making processes within the
board; the required skills of the board; and the means by which the board is held responsible to shareholders
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 33 | Page
(IOPS, 2008/9). Chow (2005) argued that a firm's various CG practices shape its behavior and eventually affect
its stock market value. Similarly, Shleifer and Vishny (1997) and Watson (2007) affirm that governance has an
association with increased investor trust, decline in misappropriation by management,reduces costs of
overregulation and improves the economic growth of countries. There are however, no universally accepted CG
principals that serve to protect and promote shareholders’ assets hence governance principles used across
countries vary (Donaldson et.al, 2001). Some common elements underlie good CG including accountability,
transparency, rule of law, inclusivity and disclosure(Bhasin, 2013).
The issue of governance is pertinent to pension schemes (Palacios, 2001). It was brought to attention
following a wave of regional market crisis and large corporate failures particularlythe Asian Financial crisis of
late 1997, the collapse of the Enron Corporation in the US and the Swissfirst affair involving Pensionskassen in
Switzerland (Stewart & Yermo, 2008) that resulted in loss of pension assets. Governance regulation serves to
reduce possible agency problems of conflict of interest and agency costs that affect negatively the security of
pensions. Agency theory depicts management of firms as agents whose interest may diverge from those of the
principals, the shareholders where both parties are utility maximizers (Jensen & Meckling, 1976). Thus, given
the choice between the two alternatives, the agent or the principal will choose the option that increases his or her
individual utility. The primary focus of good CG is in the execution of contracts that get the best out of business
performance and lessen risk (Eisenhardt, 1989). The theory thus helps minimize agents from acting
inappropriately as well as motivates them to act in the interest of shareholders (David &Impavido, 2003).
It is argued that CG mechanisms and management control effectiveness play significant roles in
enhancing financial performance of pension funds.The board of directors has a responsibility of ensuring the
implementation of good CG practices (Shleifer &Vishny, 1997). Malik et al., (2016) argues that the relationship
between board size and firm value is still a fundamental issue for research aslarge and small board size, both
have their own advantages and criticisms. The board size affects the process of monitoring, decision making and
disclosure, important functions of the board of directors. Empirical research findingshowever, are mixed across
countries and industries, and still it is a matter of research (Yang et al., 2009). The independence of the board in
a firm is evaluated by finding the presence of non-executive directors on board (Gallo, 2005). Theyare believed
to be more effective monitors who play a significant role in value creation for the firm (Butt & Hassan, 2009).
They reduce the conflict of interest between contracting parties and they are expected to act at the best
shareholders’ interest.
Agency theory’s theorists argue that the larger number of non-executives in the board might effectively
monitor the top management and protect the shareholders and other stakeholders by ensuring that there is no
collusion with top managers to expropriate minorities’ wealth. The role of audit committee isto protect the rights
of shareholdersand enhance the financial and managerial capabilities of firms (Aanu et al., 2014; Epps &
Cereola, 2008).Boards rely on their audit committees to offer effective oversight of the annual auditing process.
Audit committees also oversee the system of internal controls and ensure that the company is compliant with
laws and regulations. The CEO duality means that the CEO is also holding a position as a Chairman of the board
of directors. Yang and Zhao (2014) notes that the separation of post of CEO and the Chairman of the Board is
one of the most debatable CG issues in recent years. Strier (2005) identified the CEO duality as a fundamental
conflict of interest, having the CEO lead the group that is monitoring his or her performance. No law however,
prohibits firms from having one person perform both duties. Empirical study findings on theeffect of CEO
duality on firm performance are mixed and inconclusive (Dalton et al. 1998;Wellalage& Locke, 2011).As a
result of these inconclusive findings regarding the impact of CG on firm performance, it is worthy to study these
variables in new market environment.
Recent developments in the pension industry have made policymakers in many countries attempt to
address flaws in governance through varying combination of legal and regulatory instruments, voluntary codes
and principals. The Sarbanes Oxley Act (SOX) of 2002 in the USand the Retirement Benefit Authority(RBA)
Act Cap 197 of 1997in Kenya were enacted. Besides, the Cadbury Code in the UK, Cromme Code in Germany
and the Code of CG in Pakistan and the Mwongozo Code of Governance for State Corporations in Kenya were
developed and implemented(Kamran & Shah, 2014; PSC & SCAC, 2015). The SOX Act, together with later
regulations main objective was to protect investors from false financial reporting and fraudulent financial
practicesand improve the accuracy of their revelations.The RBA Act, Cap 197, put in place CG requirements to
be fulfilled by schemes to ensure their smooth functioning (RBA, 2014). CG weaknesses nonetheless, persist
worldwide leading to under-performance of a number of pension funds. This raises the question of why
governance reforms are not safeguarding pension benefits. Could there be other factors coming into play?
Moreover, the dynamics and development of CG in developed economies is different from those in developing
countries. Variations in basic legal systems, political stability, market size, corporate ownership and the nature
of individual financial systems create the differences in the institutional arrangements between the two (Gul
&Tsui, 2004). Evidence of the impact of CG on pension performance in developing countries is still relatively
scarce hence the need for further research.
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 34 | Page
1.1.2. Investment Strategy
Investment strategy refers to a set of rules or guidelines that help investors' select investment portfolios
based on investment objectives and tradeoff between risk and return (Bilaus, 2010). It plays a key role in
portfolio management. Pension fund management is part of the very large global investment management
industry with their assets forming a significant part (Tonks, 2006). The investment process nevertheless, faces
many challenges including inflation risks, market, credit, and solvency risks, governance and agency risks, legal
and regulatory risks that lead to lower retirement income (Obermann, 2005). This is compounded by the fact
that pension schemes unlike other saving vehicles, are long-term in nature under which the savings cannot be
accessed until retirement. Managing these risks is therefore key for ensuring their sustainability. Consequently,
it is critical that the investment function is managed responsibly.
The type of pension schemes have gradually undergone reform since the early 1980s,moving from
defined benefit (DB) systems and unfunded pay-as-you-go systems (PAYG) to arrangements in which the
provision of pensions is backed by assets in schemes, which increasingly links retirement incomes to the
performance of these assets (Rudolph et al., 2010). This results in participants being exposed to the uncertainties
of investment markets that determine the level of benefits that they will ultimately receive. To mitigate the risks,
it is argued that there will be need to employ investment management strategies that will provide income
replacement at retirement age. In addition, the aging populations has led to the explosion of the liabilities of
public PAYG schemes, the implicit pension debt. The movement to funded schemes has therefore been
motivated by governments seeking to lessen the fiscal impact of aging populations, diversify the sources of
retirement income and mitigate increasingly intergenerational transfers.
The investment strategy is a major approach that serves to manage and control these risks and expected
returns (Raz, 2005). The type of strategy adopted depends on short and long-term goals, risks involved along
with levels of variability that are acceptable. Pension schemes are unique and therefore there is no one solution
that fits all. The strategies employed include asset allocation; active or passive fund management;
diversification; limitations on portfolio allocation; market timing; indexing; as well as international investment
(Urwin, 2010). The principal theory guiding the selection of assets of a portfolio is the Modern Portfolio Theory
(MPT) of Markowitz (1952). The theory is concerned with risk and return and is based on the mean-variance
efficiency for assets allocation postulating that diversifiability of idiosyncratic risk is associated with the
expected rates of return on securities through optimal portfolio selection. It provides for a mechanism to find the
optimal combination of securities having minimum variance.Tonks (2006) notes that there is a link between
performance of pension funds and the investment strategy used.
To enhance portfolio management, the OECD issued Guidelines on Pension Fund Asset Management
that sets out a basic framework for the regulation of pension fund investment. The guidelines entail setting
retirement income objectives and prudential principles; prudent person standards; investment policy; portfolio
limits; and valuation criteria of pension assets (OECD, 2006). Sharpe (1992) established that asset allocation
accounts for a large part of the variability in the return on a typical investor's portfolio.Elton,GruberandBlake
(1996) are of the view that it is possible to outperform the S&P 500. However other researchers such as Sharpe
(1991) and Ippolito&Turner (1987) found that actively managed funds on average underperform the index, net
the costs. Bogle (2002) shows that the index performs better than the active managed portfolios in most
cases.The results support the Efficient Markets Hypothesis (Fama, 1969), which states that asset’s prices fully
reflect existing information, making it is impossible to beat the market.The mixed findings create a need for
further research.Locally, empirical literature is limited on effects of pension fund investment strategy on
performance.
1.1.3. Macroeconomic Factors
Macroeconomic factors refer to influential financial, natural, or geopolitical events that broadly affects
a regional or national economy, affecting a large population and are uncontrollable and beyond the direct
influence and control of the organisation (Brinson et al., 2009). The factors relate to the state of the economy
and government policy and include Gross Domestic Product, changes in unemployment, interest and inflation
rates, the money supply, natural disasters such as earthquakes as well as fear of civil or international war
(Sharan, 2009). The indicators are meticulously observed by investors (Chelangat, 2014). The nature of
financial decisions (investment, financing, working capital and dividend), whose ultimate goal is wealth
maximization, varies from one firm to the other. The reason is that these decisions are influenced by the
prevailing macroeconomic factors among others. Khaparde (2014) concurs that investors select assets in a
portfolio based on these factors to achieve portfolio performance that is greater than the market’s overall return.
Kahraman (2011) affirms that variations on the investment decisions of individual firms are attributed to
macroeconomic factors (Liu & Pang, 2009).
Stephen Ross (1976) formulated the Arbitrage Pricing Theory (APT) which states that there is a
relationship between financial performance of firms and the macroeconomic variables. The theory offers a
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 35 | Page
multifactor pricing model for securities and states that the return of securities is a linear function of the
macroeconomic variables. Studies by Mookerjee and Yu (1997); Kwon and Shin (1999); Humpe and
Macmillian (2007); Bodie et al. (2008); and Pilinkus (2010) in developed countries and EME established that
real GDP, industrial production, lagged inflation and interest rate had an impact on stock performance. Olweny
and Omondi (2011) and Ochieng and Oriwo (2012) found a significant association between firm performance
and the Nairobi Securities Exchange (NSE) index. Locally, limited empirical literature is available.
1.1.4. Financial Performance of Pension Funds
The financial Performance of pension funds refers to a methodology used to determine the level of
attainment of the financial objectives of a firm for a particular timeframe(Grabenwarter&Weidig (2005).
Similarly, Ijaz and Faizan (2016) describe financial performance as the extent to which an organization’s overall
financial health over a period of time is measured. It provides information on the business sector outcomes and
results for its shareholders to help them in decision making as it indicates how well an entity is utilizing its
resources to maximize the shareholders wealth and profitability. Walker and Iglesias (2007) notes that the
purpose of measuring portfolio performance is to determine whether portfolio managers add value compared to
passive investment strategies represented by feasible and well diversified benchmarks. Fama’s (1991) Efficient
Markets Hypothesis however, postulates that it is difficult for managers to add value, as asset prices fully reflect
all available information hence it is impossible to beat the market consistently on a risk-adjusted basis since.
Performance measurement is key in the effective management of an organization and in enhancement
of its processes (Carton, 2004). Kuratko and Morris (2003) observe that business environments have
uncertainties that make it trickier to predict and manage factors that can control their performance. Cheema and
Din (2013) state that performance can be used to understand CG practices and their contribution in enhancing
the total value of companies. They note that pension schemes are under scrutiny by various stakeholders
including policymakers, investors and fund trustees with and subject to performance expectations. Performance
information enables stakeholders to measure and compare the efficiency of the investment.
A complete evaluation of a firm's financial performance takes into account various kinds of measures
but most commonly used in the field of finance and statistical inference are financial ratios particularly,
liquidity, solvency, profitability and valuation ratios, trend analysis as well as market value, average annual
return and standard deviation (Tapia, 2008a,b; Ijaz &Faizan, 2016). Ratios express the numerical relationship
between two or more variables and play an important role in determining the financial strengths and weaknesses
of a firm relative to that of other firms in the same industry. The traditional approach to pension funds’
performance evaluation uses riskadjustedperformancemeasures including Sharpe’s, Sortino’s, Treynor’s ratios
which quantify the ability of pension fund managers to deliver an active management risk premium, with respect
to benchmarks. The ratios evaluate fund returns but integrate measure of risk. The study will find the Sharpe’s
ratio for the various pension funds under study.
Theratioshowshowwellthereturnofaninvestmentcompensatesfortheriskinvestorstake.ThehighertheSharperatiothe
betteritcompensatesforrisk.The grading threshold of the ratio will be i) <1 – Not good; ii) 1-1.99 – OK; iii) 2-
2.99 – Really good; and iv) >3 – Exceptional(Sharpe, 1966), where Return on assets/portfolio = Net Income ÷
Average total assets.
Sharpe’s ratio = Return of a portfolio (RP)– Risk free rate (RF)
Standard deviation of portfolio’s excess return(P)
A major drawback with risk-adjusted performance measures is that they exacerbate the herding behaviour
around the mean manager. A further potential problem is that the benchmark used for comparison may be
inappropriate such as the Market index (Fama & French, 1996).
1.1.5. Pension Schemes in Kenya
A Pension scheme is a retirement income plan that is a legally binding contract with a retirement
objective with the benefits being provided upon retirement. The pension plans may offer additional benefits,
such as disability, sickness, and survivors' benefits (OECD,2002). The OECD using the multi-pillar approach
identified three types of pension schemes:the First pillar, publicly managed pension schemes with defined
benefits and pay-as-you-go finance, based on a payroll tax; the second pillar, privately managed pension
schemes that are provided as part of an employment contract; and the third pillar, personal pension plans in the
form of saving and annuity schemes. Private schemes are managed by fund managers and insurance companies.
Pension schemes may further be classified based on two approaches: functional and institutional resulting to
plans being either public or private; occupational or personal; Defined Benefit (DB) or Defined Contribution
(DC); funded or unfunded.
In Kenya, classification of pension schemes is based on the multi-pillar approach: Pillar I- the Public
Service Pension Scheme and the National Social Security fund (NSSF); Pillar II- Occupational pension
schemes; and Pillar III-Individual pension plans. A total of 1,268 occupational pension plans and 34 individual
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 36 | Page
pension schemes exist. Majority of the pension plans have joined 26 Umbrella Retirement Benefits schemes,
which pool companies that find it is not financially feasible to create their own pension schemes. Prior to 1997,
the industry was largely unregulated and lacked a comprehensive policy framework for fostering sustainable
social protection programmes. In 1997 the government restructured the sector to address existing and emerging
issues by enacting the Retirement Benefit Authority (RBA) Act Cap 197 that established the RBA whose main
function was to oversee the growth and development of the retirement benefits schemes and sector in the
country. The financial performance of pension schemes in Kenya nevertheless continued to experience major
challenges of operational malpractices, misappropriation of scheme funds and lack of transparency, resulting in
declined pension assets. The Kenya Medical Research Institute (KEMRI) pension fund for instance lost KS 295
million held in trust account (Naftali, 2005), whilethe Kenya Ports Authority (KPA) Retirement Benefits
Scheme lost KS 700 million in 2018. The events were aggravated by poor performance of the economy. Limited
empirical literature is available on the above factors on pension performance,hence the need for further research.
1.2 Research Problem
Regional market crisis and large corporate failures in the past two decades have brought to attention the
importance of governance. The 1997 Asian Financial Crisis of the “tiger economies" for instance saw their
capital markets and currencies lose 70% of their values (Kuepper, 2019). The Global Financial Crisis of 2008 as
well resulted in the great recession (Amadeo, 2019) leading to an estimated loss of US $5.4 trillion or about
20% of the value of pension assets in OECD countries (Antolín & Stewart, 2009). The accounting scandals
resulted to a breach of the governance and compliance practices leading to failure of the firms. These events
reveal the severity of the agency problem that arises between managers, shareholders and stakeholders that has
threatened to erode contributions that pension funds make to the world economies (OECD, 2008; Rudolph et al.,
2010). This has exacerbated the threat of pension funds failing to provide retirement benefits (Besley & Prat,
2005). Accordingly, it has put pressure on governments and pension funds to initiate CG reforms to address the
problem.
Similarly, challenges of operational malpractices, misappropriation of scheme funds and lack of
transparency were witnessed in Kenya, aggravated by poor performance of the economy. The Kenya Medical
Research Institute (KEMRI) pension fund lost KS 295 million held in trust account (Naftali, 2005) while the
Kenya Ports Authority (KPA) Retirement Benefits Scheme lost KS 700 million through illegal purchase of
assets. The challenges persisted despite the enactment of the RBA Act Cap 197 in 1997 that established the
RBA that was to provide oversight on the growth and development of the pension sector. A review of existing
literature suggests that only a limited number of studies were carried out to evaluate the impact of CG on
pension performance (Palacios, 2002). This has led to a fairly fragmented and developing literature in the area,
giving a lack of clarity over many concepts. The significance of this study is driven from the importance of the
pension sector itself contributing 13% of the GDP.
It is argued that good pension governance is a critical aspect of a well-functioning retirement system as
it is posited to determine investment performance and hence security of retirement benefits. Studies by Yang
and Mitchell (2005), Manuel and Andreas (2008) and Clark and Urwin (2008) found positive associations
between good CG and firm financial performance. Other studies (Daines& Klausner, 2001; Coles, et al., 2008;
Bhagat & Black, 2002) found mixed and inconclusive results on the association between CG and pension fund
financial performance. Lack of unanimity continues to make the subject a current issue requiring further
research to enable a better understanding of the relationship between the study variables and the Agency theory.
The theory expounds on the association between the principal, who employs another party, the agent to work on
its behalf in an organisation. The agent may not act in the principal’s best wishes (Jensen & Meckling, 1976).
Due to the varied level of development of capital markets, there is a likelihood of getting different results when
studies are carried out in developing economies. This brings to the fore the influence of other factors such as
investment strategy and macroeconomic variables. Could they be influencing the above relationship.
Investment strategy is another major factor postulated to influence performance of pension funds. It is
an approach to manage and control risksthrough a blend of strategies. It guides an investor’s selection of
investable assets through a tradeoff between risk and returnas guided by the Markowitz’s (1952) Portfolio
Theory. The theory provides a framework within which to make sensible asset management and allocation
decisions. It proposes that all investors are risk averse and that risk can be reduced by combining dissimilar
financial assets to form a diversified investment portfolio. Management of risks was only made wide open by
the recent Global economic turmoil’s that led to loses in value of pension assets worldwide. Studies in the
developed and EME by Mitchell and Hsin (1997); Iglesias and Palacios (2000); and Davis and Hu (2008)
established that there is a relationship between good CG, investment approaches and pension performance.
Locally however, there is limited empirical literature on the effect of investment strategy on pension
performance, leaving a lacuna.
Macroeconomic factors as well are critical consideration by institutional investors when it comes to
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 37 | Page
assets under management (Brinson et al., 1991). Flanery and Protopapadakis (2002) and Singh (2010) identified
macroeconomic factors together with legal and regulatory environment as the variables that influence financial
performance of stocks. The Arbitrage Pricing Theory a multi-factor asset pricing model holds that an
asset’s returns can be forecasted using the linear relationship of an asset’s expected returns and
the macroeconomic factors implying that the financial performance of pension funds’, major investors in stocks
in capital markets will be affected by prevailing systematic factors. Studies in the developed world and EME by
Fama and French (1989); Mookerjee and Yu (1997); and Kwon and Shin, (1999) affirm that there is a link
between stock market return and systemic factors particularly, GDP, interest and inflation rates. Other studies
however, found mixed and inconclusive results on the effect of systemic factors on stock returns (Chan et al.,
1998; Flannery & Protopapadakis, 2002). The mixed finding provides a need for further research to better
understand the relationship between the variables and the APT. Locally, studies on the subject have been few
resulting in limited empirical information. Those done tended to focus on other sectors and used different
methodologies.Olweny and Omondi (2011), Ochieng and Oriwo (2012) and Osoro (2015) investigated and
established that macroeconomic variables influenced stock returns and growth of capital markets in Kenya.
Their focus however, was not on pension funds.
The above reviewed empirical literature reveals a number of research gaps. There is little insight
relating the performance of pension funds to CG, investment strategy, and macroeconomic variables as multiple
factors. There is lack of consensus on why similar CG practices in the developed world produced conflicting
and sometimes inconclusive results. Most studies did not take into account the interaction between intervening,
moderating and governance structures on pension performance. The studies too neverused the multi-equation
approachto assess the influence of multiple factors on pension performance. The focus of this study, to
investigate the joint effect of the CG, investment strategy and macroeconomic factors on pension performance.
This will help explore the causal relation amongst these factors and shade light on the nature of the relationship
from a developing country’s perspective. The study thus endeavours to seek solutions to the key study question:
How have CG, investment strategy and macroeconomic factors influenced the financial performance of pension
funds in Kenya?
1.3 Research Objectives
The main purpose of the study is to assess the level of association between governance, investment strategy,
macroeconomic factors and financial performance of pension funds and in the country. Specifically, it will seek
to:
i) Examine the influence of corporate governance on financial performance of pension funds.
ii) Establish the effect of investment strategy on the relationship between CG and financial performance of
pension funds.
iii) Establish the effect of macroeconomic variables on the relationship between CG and financial
performance of pension funds.
iv) Evaluate the joint effect of CG, investment strategy and macroeconomic variables on the financial
performance of pension funds.
1.4 Value of the Study
The study outcomes will provide empirical evidence on descriptive statistics on the association
between CG, investment strategy, systemic factors and financial performance of pension funds. The findings
will link these factors and pension performance in an integrated manner and extend the CG and pension
performance discussion. In addition, it will provide evidence from a developing country’s perspective on the
application of the theories anchoring the study. Both theory and empirical findings will contribute to our
understanding of the interplay between research variables and will provide useful knowledge to practitioners,
policy makers, trustees and plan members to make sound and effective strategic decisions to achieve superior
pension performance. In addition, the study will help bridge the gap between research and practice. Research-
based knowledge will lead to greater organizational effectiveness.
CG and the risk management were at the heart of the debate on the 2007-2008 financial meltdown and
the large corporate failures. The study therefore will help identify drivers of effective CG and unearth factors
that are key to the investment process. The findings will assist investment managers; plan members and
beneficiaries make sound and informed investment decisions in asset allocation, portfolio construction and risk
management to improve financial performance of pension schemes. Moreover, theresearch findings will be
relevant also to the regulators (RBA &CMA) and market participants (NSE). The regulators can use the findings
to guide the regulation process. They can also be used to examine critical areas of CG and to formulate
necessary policies as guiding frameworks for CG.
Empirical studies on CG in Kenya are limited and therefore it is conceivable that research results will
help build academic knowledge in the investment management, systemic factors and pension performance. This
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 38 | Page
will guide academicians and research institutions in the development of acceptable CG models relevant in the
context of a developing economy. Hess and Impavido (2003) opine that knowledge of the CG theory supports
the adoption of good CG practices to reduce agency glitches in pension schemes. The applicability of the
research theories and models developed in developed countries to that in the context of a developing country
such as Kenya will be examined as there exist differences in political, legal, economic, social and cultural
settings. The successful use of these theories in this study will contribute towards providing examples of the
interpretation of studies from developing countries perspective. Finally, the study will bring to light the
importance of the sector to key stakeholders.
II. Literature Review 2.1. Introduction
One area in which dominant ideologies of finance of pension systems converge today is the agreement on the
need to ensure financial solvency in larger economic and fiscal matters affecting pension systems. The chapter
will review both empirical and theoretical literature on the association between CG, investment strategy,
systemic factors and financial performance of pension funds as outlined below.
2.2. Theoretical Foundation of the Study
The founding theory for this research will be the Agency Theory. The study will however be backed three other
theories: The Modern Portfolio Theory (MPT), the Stakeholders Theory (SHT) and the Arbitrage Pricing
Theory (APT).
2.2.1. The Agency Theory
The Agency theory expounds on the association between the principal, who employs another party, the
agent to work on its behalf in an organisation (Jensen &Meckling’s, 1976). The agent may not act in the
principal’s best wishes. The relationship between the two results from the separation of ownership and control
which requires protection of shareholders’ interests, minimise agency costs and align principal-agents interest
(Demsetz& Lehn, 1985). The theory states that rational actors, both agents and principals seek to maximise their
individual utility with the least possible expenditure. Given the choice between the two alternatives, the agent or
the principal will choose the option that increases his or her individual utility. It is nonetheless difficult for the
principals to know ex-ante which agents will self-aggrandise, and so it is prudent for the them to limit potential
losses to their utility (Williamson, 1985). The theory therefore aims at reducing agency costs incurred by the
principal by imposing internal controls that keep the self-serving agent’s behaviour in check. This harmonizes
the interests of the managers and the shareholders to maximize company value (Maher & Andersson, 1999). To
protect shareholders interests, minimise agency costs and ensure principal-agents interest alignment, agency
theorists prescribe various governance mechanisms (alternative executive compensation schemes and
governance structures (Demsetz& Lehn, 1985).
The financial incentive schemes provide rewards and punishments that are aimed at aligning principal
agents’ interests. For governance structures, boards of directors keep potential self-serving managers on check
by performing audits and performance evaluations. Outside (non-management) board leadership and
membership are desirable to ensure that proper management oversight occurs. The study will examine the
relationship between pension performance and governance structures of board size and diversity, CEO duality,
presence of director(s) from institutional investors and stakeholder engagement.
Critiques of the Agency theory have however grown with time. Donaldson (1990) and Aguilera et al.
(2008) point out its narrow nature that makes it difficult to compare and explain CG practices across different
institutional and national context. Equally, Shapiro (2005) critiqued the theory for its narrow analytical focus in
the context of shareholders as the only ones with interests in the listed firms. Doucouliagos (1994) argued that
labeling all motivation as self-serving does not explain the complexity of human nature. The Stewardship theory
argues that shareholder interests are maximised by shared incumbency of these roles. Consequently, managers
are not motivated by individual goals but rather are stewards whose motives are aligned by the objectives of the
principals (Donaldson & Davis, 1991). A steward’s behaviour is such that pro-organisation, collective
behaviours have higher utility than individualistic self-serving behaviours hence he/shewill not depart from the
interests of his or her organisation. The Agency theory nonetheless is justified in the study as it provides direct
relationship between governance and pension performance. It offers a useful way to explain relationships where
parties interest are at odds and can be brought into alignment through proper monitoring and well-planned
compensation system.
2.2.2. Stakeholder Theory
The Stakeholder Theory (SHT) of CG is about identifying groups who are participants in the
corporation that need to be managed (Freeman, 1984 and Aguinis &Glavas, 2011). Stakeholders of a firm
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 39 | Page
comprise individuals and constituencies with different interests and values that contribute to its wealth creating
capacity and activities and who are therefore its potential beneficiaries and or its risk bearers (Preston &
Donaldson, 1995; Post et al., 2002).The theory states that, apart from the shareholders, the achievement of a
firm has a correlation with other stakeholders who have interest in the firm. It suggests that a wider constituency
interests judge firm performance. Since companies are liable to a bigger group of stakeholders other than
shareholders, Mayer (1996) is of the view that they should be managed to serve public interest. The theory
legitimizes its value as an effective means to improve efficiency and economic success. The theory has both
normative instrumental implications (Preston& Donaldson, 1995); Jones & Wicks, 1999). Normative
implications mean it has a moral/ethical responsibility to meet the legitimate claims of all stakeholders.
Instrumental means it has a profit/wealth enhancing obligation - a means to maximize organizational wealth.
Stakeholders should therefore participate in corporate decision-making process as a way of enhancing efficiency
to achieve specific goals (Kelly & Parkinson, 1998). The theory therefore is primarily concerned with how CG
practices promote the interests of both the shareholders and other stakeholders (Williamson (1985). Health and
Norman (2004) nonetheless, critiques the SHT by observing that poor company performance may be justified by
management through the use of stakeholder reasons. Blair (1995) too, notes that there is a challenge of
attainingfirms’ wider objectives.
Pension performance is postulated to have a relationship with the interests of stakeholders. Limited
research has however, been undertaken on how to integrate interests of all stakeholders into the scheme’s
decision making and management processes. Studies already done have investigated whether or not companies
that perform well on measures of social performance also perform well on economic measures (Jones, & Wicks,
1999). Others examined the firm’s role to satisfy a wider set of stakeholders, not simply the shareowners
(Alkhafaji, 1989). In addition, there are those that focused on the perception of the board members regarding
their stakeholders or corporate social responsibility (CSR) orientation (Agle et al., 1999; Wieland, 2005) or on
the representation of stakeholders on the board of directors (Hillman et al., 2001). Their findings indicate that
stakeholder engagement enhances firm performance.Similarly, Demsetzand Lehn (1985) and Wallace and
Cravens (1983) established that large US public firms with nomination committees (audit committees,
shareholder relation committees)that protect shareholder rights performed better infinancial performance than
companies without.Locally, limited empirical studies on the subjectare available.
2.2.3. Modern Portfolio Theory
The Markowitz (1952) Modern Portfolio Theory (MPT), the Efficient Frontier is an investment theory
that provides a framework within which to make sensible asset management and allocation decisions. The
theory postulates two main concepts: i) all investors have a basic objective of attaining maximum returns for any
level of risk, ii) risk can be reduced by combining dissimilar financial assets to form a diversified investment
portfolio. Risk is categorized into systemic and un-systemic risk.Systematic risks are those inherent in the
capital market whereas unsystematic risks are associated with each particular stock as they are company-specific
events and risks (Sharpe, 1964; Lintner, 1965). The later are lowered by diversification. The Efficient Frontier,
as proposed by the theory is a graphical representation of all possible combinations of risky securities based on
the best level of return given a particular level of risk. Investors select their preferred portfolios based on their
specific risk predisposition. The theory functions on assumption of investors being risk averse, hence they
expect to be rewarded for taking additional risk; are rational; and have access to comparable information.
Lately, scholars have critiqued the theory. Studies by Haugen and Heins (1975) and Murphy (1977) established
that risk-reward relationship was far weaker than expected. Moreover, the theory’s assumptions are incorrect.
Gregory (2002) reports that behavioural economists have shown that not all investors act rationally. In addition,
all investors are not equally informed, as the market is asymmetrical with information due to insider trading.
The theory will be used to evaluate the investment strategies that will help construct portfolios that maximize
expected returns and minimize investment risk. Despite these weaknesses, the theory is still applicable in
investment management.
Pension plans manage assets that are used to provide workers with a flow of income during their
retirement years. Because the plans control the largest pool of capital in the world, asset managers need to be
aware of the goals and challenges of managing these plans. Watson (2011) in a study of 13 developed countries,
showed that private and public pension plan assets were over US $26 trillion, averaging 76% of gross domestic
product (GDP). The investment process of the pension assets is guided by the MPT that rests on the foundation
that risk-averse investors can construct portfolios to maximize expected returns based on a certain level of
market risk.
A review of studies on the performance of investment funds have revealed mixed results. Studies have
been undertaken to evaluate the performance of pension funds on the basis of the economic trade-off between
portfolio risk and return. Blake, Lehmann and Timmermann (1999) analyzed a data set on UK pension funds.
Their main finding was that strategic asset allocation accounts for most of the ex post variation of UK pension
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 40 | Page
funds’ returns. Other studies established that the vast majority of funds had negative market-timing estimates
(Coggin et al., 1993; Daniel, et al. 1997; Blake et al., 1999). Oppolito (1989) looked at mutual fund data and
found evidence that is consistent with optimal trading in efficient markets. In contrast, Grinblatt and Titman
(1989) looked at mutual fund performance and tests indicated that the risk-adjusted gross returns of some funds
were significantly positive. They concluded that risk-adjusted returns in the mutual fund industry, net of fees
and expenses, are comparable to returns available in index funds. The findings show that there are those that
support market efficiency as well as those that reject it. The later are of the view that investors can apply the
MPT to attain an optimal risky portfolio that is fully diversified to achieve a higher return than investing in an
index portfolio.This makes the theory relevant to the study. The mixed findings mean that the is need for further
research.
2.2.4. The Arbitrage Pricing Theory
The Arbitrage Pricing Theory (APT) developed by Ross (1976) postulates that there is an association
between expected return of a security and a set of systematic factors that affect the assets risks. The theory
offers a multi-factor pricing model for securities. The author affirms that diversification of portfolios does not
eliminate risks completely as there are economic forces that still influence stock returns. Studies by Chen, 1986;
Roll & Ross, 1980 on the model shows that factors such as GDP, changes in inflation and interest rates affect
expected stock return.
The main weakness of the theory is on its generality. It fails to explain the theoretical reasons for
selecting identified systemic factors as well as their number (Huberman, 2005). Roll (1977) argues that it is
difficult to test the theory, as the precise configuration of the market portfolio is not known. Estimation of the
model also faces certain challenges relating to methodologies used. Cheng (1996), Günsel and Çukur (2007)
established that the number of independent variables used influences the model. In the later two cases, it was
found that the applicability of the APT in establishing asset returns may still be valid. The theory will be used to
interrogate the association between pension financial performance, CG and systemic factors.
2.3. Empirical Review
The section presents empirical literature outlining the relationship between CG, investment strategy, systemic
factors and financial performance of pension funds. The studies are relevant as they provide the empirical
relationship of the variables and the applicability of the theories.
1.1.1. Corporate Governance and Pension Performance
Existing empirical literature on CG is mainly from US and OECD firms (Maher & Andersson, 2000).
Research finding showed that the financial performance of firms was influenced by the level of shareholder
rights and the competence of existing court systems (Gompers et al., 2001; La Porta, et al., 2001; Lombardo
&Pagamo, 1998). In particular, they established that enhanced shareholders’ rights resulted in higher financial
performance of firms. Besley and Prat (2003), Mitchell and Yang (2005), and Manuel and Andreas (2008) found
positive relationship between good CG and pension performance. Wagner et al. (1998) found that the
probability of firms going under declined with boards controlled by outside directors. Zahra and Pearce (1989)
aver that outsiders tend to be objective, unbiased and independent.
Mixed and sometimes inconclusive results on the relations between CG and firm performance were
also found by scholars such as Daines and Klausner, 2001 (examined takeover defenses), Larcker, et al. (2007)
(examined board and ownership variables) and Coles, et al. (2008) (considered board size). Clarke
(2009)observed that CG systems failed to prevent financial crisis and corporate collapses across different
economies. Heracleous (2001) reports that researchers failed to find any convincing connection between the best
practices in CG and organizational performance. A possible explanation for these results is that there could be
other factors influencing the above. Renders et al. (2010) attribute it to the differing and limitation of methods of
measuring CG and econometric problems.
Studies on CG of pension funds in Kenya are in the early stages of development and have tended to
focus on different sectors. Available empirical evidence is therefore indirect and not related to pension funds.
Moreover, different methodologies and variables were used. Mutegi (2014) established that CG structures of
occupational retirement benefit schemes in Kenya had a correlation with the financial performance of pension
plans. Njuguna (2011) found that good CG practices had a positive correlation with pension regulations,
leadership and growth of schemes. None of these studies examined the influence of other factors on the above
relationship. Ongore and Kobonyo (2011) assessed the relationship between financial performance of NSE
listed firms and governance. They established significant relationships between ownership concentration and
profitability of firms. Miring’u (2011) showed that the performance of board members significantly influenced
the financial performance of state firms. Lishenga (2012) assessed the effects of board meetings for CG on firm
performance and established that improved regularity of board meetings enhanced firm performance. Arising
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 41 | Page
from these findings, one notes that the focus was on firms and not pension funds. None of the studies too
assessed the effect of several factors using a multi-equation approach or a composite measure of CG on pension
performance. Further studies are thus required to establish the effect of these factors using a multi-equation
approach from a developing countries perspective.
1.1.2. Corporate Governance, Investment Strategy and Pension Performance
The effect of governance on investment decisions in institutional investors, private equity funds and
pension funds was undertaken by Khanna and Zyla (2012) in emerging markets (EME). They established that
CG was an important factor when making investment decisions and investors were prepared to pay better prices
for firms executing good CG practices compared to those poorly governed. The study however, did not
investigate the role of trustees in the investment process. In contrast, Useem and Mitchell (2008) showed that
CG has no relationship with the financial performance of investing firms. The authors however, showed that
governance, influenced the kind of investment strategy used, which had a positive relationship to the financial
performance of investments of pension funds. Thus, the financial performance of the funds’ investments is
indirectly affected by CG.
In Switzerland, Manuel and Christian (2016) investigated the relationship between CG, asset allocation
and financial performance of 139 Swiss pension plansundertaking investment opportunities. They established
that there is a direct relationship between CG and financial performance of pension plans. The relationship
however, is only slight to the category of assets selected. Ambachtsheer, Capelle and Scheibelhut (1998)
evaluated the impact of quality of governance structures on financial performance of pension funds undertaking
investment opportunities. Their findings showed that the relationship was positive.In Poland, Jackowicz and
Kowalewski (2012) showed that there is a positive correlation between the number of outsiders on trustee
boards, the level of education, and the market values of the funds. A review of the studies indicates that
identifying and understanding the persistence of the poor performance of some fund managers is an important
issue despite the fact that the average disguises the fact that some fund managers perform well, and others
perform poorly. None of the studies were carried out in developing countries. Furthermore, the level of capital
market development varies greatly between the developed and developing countries. This may affect the
outcome of the study. Studies carried out too did not take into account the interaction of multiple factors. It is
against this backdrop that this study is undertaken to fill the gap.
1.1.3. Corporate Governance, Macroeconomic factors and Pension Performance
Most of the evidence available on studies examining the sources of return variation is indirect and not
necessarily related to pension funds but to securities that pension funds invest in. Research in developed
countries and EME (Chen,1991, Black et al., 1997; Humpe &Macmillian,2007; Mukherjee & Yu, 1997; Kwon
& Shin 1999) showed that real GNP, industrial production, lagged inflation and interest rate influenced stock
performance. Muhammad and Rasheed (2002) evaluated the influence of interest rates on stock return for firms
in Pakistan, India, Bangladesh and Sri Lanka using monthly data from 1994 to 2000. Their findings indicated a
positive relationship between the two variables for firms in Bangladesh and Sri Lanka only. No relationship was
however, found for companies in India and Pakistan.
In another study involving the Bombay Stock Exchange (BSE) Sensex, Singh (2010) assessed the
impact of exchange rates, industrial production, and wholesale price index on stock return from 1994/95 to
2008/09. The results found were mixed. The three factors had a positive relationship with stock return.
However, when the Granger causality test was used to evaluate the findings, index of industrial production was
the only factor having bilateral causal relationship with BSE Sensex. The author concluded that in the Indian
Capital Market asset’s prices fully reflect existing information on exchange and inflation rates. In the Kenyan
context, studies by Olweny and Omondi (2011) and Ochieng and Oriwo (2012) found a positive relationship
between the Nairobi Securities Exchange All Share Index (NASI), the firm’s financial performance, foreign
exchange rate, interest rate and inflation rate. Wanjiku (2012) as well found that pension performance was
heavily influenced by selected macroeconomic variables. She concluded that in the Kenyan Capital Market,
asset prices do not fully reflect existing information. There is therefore need to monitor macroeconomic
environment since these changes affect security returns. A review of the existing literature nevertheless reveals
that none of the studies investigated used a multifactor model to evaluate the impact of CG, macroeconomic
variables and investment strategy on financial performance of pension funds.
1.1.4. Empirical evidence on the joint effect of CG, investment strategy and macroeconomic factors on
pension performance
Empirical studies focusing on the effect of multiple factors on the association between CG and pension
fund financial performance are limited both in the developed and developing countries. This is a research area
that needs attention. Previous studies on the relationship between CG and pension performance attribute the
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 42 | Page
mixed findings of inconclusiveness or contradictions to the use of two variables at a time (Uwuigbe, 2012). The
study will therefore try to address this gap by using a multifactor model to investigate the joint effect of CG,
investment strategy, and macroeconomic factors on pension performance.
2.4. A Summary of Knowledge Gaps
Table1: A Summary of Knowledge Gaps Scholar &
Year
Area of Focus Study Analysis
Model
Research Findings Research Gaps Focus of Current
Study
Corporate Governance and Pension Performance
Manual and Andreas (2008)
Evaluation of the effect of CG on
the financial performance of
pension funds in
Switzerland
Cross sectional survey
Governance practices on organization and
target setting had a significant association
with the financial
performance of pension funds
The study did not consider the effects
of intervening or moderating factors
on pension
performance
Using a multifactor model the research will
examine the impact of CG, investment
strategy and
macroeconomic factors financial performance
of pension funds
Fich and
Shivdasani (2006); Coles
et al. (2008)
Analysis of the
consequences of busy boards and
assessment of the
impact of board sizeon firm
performance
ROA Outcomes yielded
mixed findings on the relations, between CG
measures and firm
performance
Mixed and
sometimes inconclusive
findings; The
studies did not put attention on
developing
economies
Examine impact of CG
practices, investment strategy, institutional
characteristic and
macro- economic factors on pension
performance
Ongare and
Kobonyo
(2011)
The impact of CG
on firm
performance on firms listed at the
NSE where
ownership as a key variable
Survey; ROA
ROE
There were significant
relationships between
ownership concentration and
profitability of firms
The study did not
consider effects of
mediating and moderating factors
on firm
performance. Focus too was not on
pension funds
The study will take
into account the effect
of interaction of CG, mediating and
moderating factors on
pension performance
Corporate Governance, Investment Strategy and Pension Performance
Khanna &Zyla (2012)
The impact of CG on investment
decisions in
different type of institutions in
emerging
countries
Survey, ROA Governance was key when making
investment decisions
The study did not consider the effects
of mediating and
moderating factors on the relationship
between CG,
investment strategy and pension
performance
The study will take into account the
interaction of CG,
mediating and moderating factors on
pension performance
Brinson, Hood and Beebower
(1986)
Determinants of Portfolio
Performance
ROA Survey Market timing and stock selection
account for only 6%
of the variation in returns in a portfolio
whereas investment
policy accounts for 94%
The study did not investigate the
impact of CG,
mediating and moderating factors
on pension
performance
The study will take into account the
interaction of CG,
moderating factors and investment strategy on
pension performance
Corporate Governance, Macroeconomic Factors and Performance of Pension Funds
Kwon & Shin
(1999)
Impact of
macroeconomic factors on value of
securities
measured by stock prices
Survey co-
integration test & a Granger
causality test
There is an association
between Korean Stock price indices and a set
of macro-economic
variables
The studies did not
take into account the interaction of
CG, mediating
factors and macro- economic factors on
firm performance
The study will take
into account the interaction of CG and
mediating factors on
the financial performance of
pension funds
Ochieng &
Oriwo (2012)
Effect 91 day T-
bill and inflation rate on the Nairobi
Securities Exchange all share
index (NAS)
Autoregressive
distributed lag (ARD) bound
test approach
There is an association
between 91 days T-bill and inflation rate
and the NASI
The studies did not
consider the effect of CG and
Mediating factors on relationship
between
macroeconomic factors and firm
performance
The study will take
into account the interaction of CG and
mediating factors on the financial
performance of
pension funds.
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 43 | Page
2.5. Conceptual Framework
2.6. Source: Author, 2019
2.7. Hypotheses
The study tested the following hypotheses:
i) H1: CG has a significant relationship with the financial performance of pension plans.
ii) H2: Investment strategy has a significant intervening effect on the relationship between governance and
financial performance of pension plans.
iii) H3: Macroeconomic variables have a significant moderating effect on the relationship between
governance and financial performance of pension plans.
a) H3(a): GDP growth rate has a significant moderating effect on the association between CG practices and
financial performance of pension plans.
b) H3(b): Inflation rate has a significant moderating effect on the relationship between CG practices and
financial performance of pension plans.
c) H3I: Interest rate has a significant moderating effect on the relationship between CG practices and
financial performance of pension plans.
iv) H3(d): Exchange rate has a significant moderating effect on the relationship between CG practices and
fiscal performance of pension plans.
v) H4: The joint effect of CG, investment strategy and macroeconomic has significant relationship on
pension performance.
III. Research Methodology 3.1. Introduction
The section comprises a review of the research procedure that comprises the research philosophy, design,
population and sample of the study, data gathering, tests of validity and reliability as well as analysis of data.
3.2. Research Philosophy
Research philosophy refers to a set of beliefs and assumptions that guide the development of new
knowledge in a particular area (Saunders, Lewis & Thornhill, 2019). Kuhn (1962) describes it as a system of
scientists’ beliefs and agreements that enables one to understand problems and find their solutions. The
philosophy comprises assumptions that support research strategy and the methods one chooses. It encompasses
the concepts of epistemology, ontology and axiology. Epistemology is the study of knowledge acquisition and
justified beliefs (Easterby et al., 2008). It entails creation and propagation of knowledge in specific areas of
research (Gertler, 2015). Ontology concerns the overall nature of reality and specifies assumptions involved
(Gruber, 1995). Axiology refers to the role of values and ethics in research (Heron, 1996).
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 44 | Page
A research paradigm is an approach to undertake a study (Kuhn, 1962). Guba and Lincoln (1982) refers
to it as a basic set of beliefs that guide action in research. Two main paradigms exist: positivism and
phenomenological (Sekaran, 2003; Westland, 2004). The authors affirm that positivism relates to the view that
involves working with an observable single reality that can be measured and known using quantitative methods
to create law like generalizations. The generalizations help explain and predict behaviour and events in
organizations. The focus on positivism is on scientific empirical approaches designed to provide unbiased data.
It uses present theories to develop hypotheses to be tested and confirmed or refuted. In contrast,
phenomenological paradigm assumes that people differ physical phenomena as they interpreter issues (Saunders
et al., 2018). To get those multiple realities, they use qualitative methods of observation, questioning and
description (Crotty, 1998). Since the study seeks to test quantitative hypotheses, a positivistic research approach
will be used.
3.3. Research Design
Research design is described as a main plan for the for the collection, measurement, and analysis of
datato address a research problem (Zikmund, 2003). It is the global strategy selected to incorporate the various
sections of the study in a clear and systematic way to address the research problem (Trochim, 2006). It is the
overall strategy one chooses to integrate the different components of the study in a coherent and logical way to
address the research problem (Trochim, 2006). Creswell (2008) identifies three research designs: qualitative,
quantitative, or mixed methods. The study used both quantitative and qualitative research designs.The
qualitative research design of in-depth interview will be used to assess both the impact of CG structures and
investment strategies on financial performance of pension schemes. They examine about persons and the reason
behind the thinking through collection of no-numeric data.The design is more descriptive and is used to draw
inferences. It involves five methodologies: content analysis, in-depth interview, focus groups,ethnographic and
case study research. The in-depth interview involved survey questionnaires, interviews and documentation
review (Neuman, 2006). Both the CG index and investment strategy index were estimated using this method.
Quantitative research designs asses the level of association between study variables using statistical
analysis techniques (Creswell, 2013). They are classified as descriptive, correlational, quasi-experimental and
experimental research designs, observing and describing the behavior of a subject without influencing it in any
way.Descriptive research describes the characteristics of the population or phenomenon that is being studied
focusing more on the “what” of the research subject rather than the “why” aspect. It describes a subject
population’s critical variables that will provide answers to the questions of who, what, when, where, and how
related with a specific study problem (Cooper & Schindler, 2003). The design involves three methods in data
collection: observational, case study methods as well as survey research. This design is used when one wants to
define respondent characteristics, measure data trends, conduct comparisons and validate existing conditions.
Correlation studieson the other hand are where a researcher investigates associations between variables
and none of the variables are manipulated (Waters, 2017). Developmental studies evaluate changes over time.
The study used descriptive, correlational, survey and developmental quantitative research designs to assess the
relationship between financial performance of pension funds and the variables CG structures, investment
strategy, interest, exchange and inflation rates and change in Gross Domestic Product (GDP). The study wasalso
longitudinal as sample members were measured repeatedly over time. The quantitative data collected included
performancemeasurements of pension funds, NSE 20 share index, exchange, inflation and interest rates, changes
in GDP.
3.4. Population of the Study
Population of a study is described as the entire set of subjects (people, objects, events, or
measurements) that have similar characteristics (Mugenda & Mugenda, 2003). Polit and Hungler (1999) defined
it as the entirety of all the subjects that fit certain qualifications. The research population comprises 1306 public
and private pension funds registered with the RBA as at 31st December 2018 organised as either individual or
umbrella pension schemes (Appendix III and IV). The unit of analysis was each of the individual or umbrella
pension schemes or targeted fund managers from these pension schemes.
1.1.1. Sample of the Study
A sample is a subsection of a population carefully chosen to take part in the study (Brink, 1996; Polit & Hungler
1999:227). LoBiondo-Wood and Haber (1998) refers to sampling as the method of selecting part of the
population to represent the entire set of subjects.To produce results that can be generalized to the population,
random sampling method was applied. Sample size was estimated using Cochran’s sample size formula
(1963:75):
n0 = Z2pq/ e
2.
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 45 | Page
Where n0 is the sample size; Z2 is the critical value of the Normal distribution at α/2, for example Z= 1.96 for a
confidence level of 95%, α is 0.05; e is the required accuracy level; p is the sample fraction with a characteristic;
and N is the entire set of subjects. The selection of the period of study is informed by the fact that major CG
reforms were effected during that time, providing a scope to evaluate the influence of CG as well as investment
strategy and macroeconomic factors on pension fund financial performance. Size of the sample for the studywas
297 estimated:
n = Z2*N*∂p /{(N-1) * ℮
2 + (Z
2*∂2
p)} n=1.962*1306*0.5
2/{(1306-1) 0.05
2+(1.96
2*0.5
2)}
Where; N=1306, the population size; e= 0.05, margin of error; ∂p = 0.5, the standard deviation of the
population; and Z = 1.96 at 95% confidence level.
3.5. Data Collection
Data used in the study comprised both primary and secondary sources entailing time series and cross-
sectional data covering the years 1997-2018, the time when major pension regulatory reforms were undertaken
in sector. Data was derived from several sources. Quantitative data on monthly value of pension assets and their
returns was obtained from individual pension funds records, annual reports or archives. Market surveys, annual
reports and publications from the Central Bank of Kenya and the Kenya National Bureau of Statistics provided
quantitative data on GDP, inflation and foreign exchange rates while the Capital Markets Authority provided
NSE 20 share index, corporate bond and T- bill rates. Primary data comprising CG and investment strategy
index was obtained after analysis of qualitative data collected using survey questionnaires from the pension
schemes. The respondents for the questionnaires included elected members of the schemes’ trustee sponsor,
elected trustee, corporate trustee scheme administrator, scheme manager, custodian actuary and any other person
with knowledge on the institution.
3.6. Tests for Reliability and Validity
3.6.1. Tests of Reliability
Reliability is “the degree of consistency with which the instrument measures an attribute” (Polit &
Hungler 1999:255). De Vos (1998) describes it as the level to which the use of a specific research tool in
another study, yields equivalent outcomes under similar settings. Cronbach (1951) referred to it as how closely
related a set of items are as a group. All the definitions embody the concept of repeatability or replicability of
research findings. Joppe (2000) avers that the research instrument is reliable if the study findings can be
reproduced under a comparablecondition. Reliability therefore is about the precision of the actual measuring
research instrument or procedure and is estimated using Cronbach’s Alpha Coefficient. The coefficient ranges
from 0-1. If all items are not correlated, then α = 0; and, if all of the items have high covariances, with
α approaching 1, they probably measure the same underlying concept.For this study, the Test re-test approach
was used to evaluate the reliability of the two sets of questionnaires of CG and investment strategy. The
questionnaires was administered and later repeated over a period of time to management personnel of several
independent pension funds.The findings from Time 1 and 2 werethen evaluated to see if there wasany
association over time.
3.6.2. Tests of Validity
Validity is a test that measures the extent to which study scores represent what it is purported to
measure (Wren, 2006). It determines how truthful the research results are and is measured by the presence or
absence of systemic error of data (Campbell & Stanley, 1963).
3.6.3. Diagnostic tests
Model diagnostics is concerned with testing the goodness of fit of a model and, if the fit is poor,
suggesting appropriate modifications. The tests are applied to evaluate model residuals, which also
serveastests of modeladequacy. They are designed to examine the dependence (correlation) structure of a time
series. If a time series is serially uncorrelated, no linear function of the lagged variables can account for the
behavior of the current variable. They include Multicollinearity tests and the Heteroscedasticity tests.
3.6.4. Multicollinearity
Multicollinearity occurs when the explanatory variables are very highly correlated with each other in a
model. Its presence can adversely affect the regression results: i) R2 will be high but the individual coefficients
will have high standard errors; ii) The regression becomes very sensitive to small changes in the specification;
iii) The confidence intervals for the parameters will be very wide, and significance tests might therefore give
inappropriate conclusions.Detecting multicollinearity is through calculation of correlation coefficients for all
pairs of predictor variables. If the correlation coefficient, r, is exactly +1 or -1, this is called perfect
A. Investment strategy
Asset allocation,
Investment style- active/passive,
Diversification,
Market timing
Limitation on portfolio allocation
Equities as a % of the total assets
International investments of some assets
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 46 | Page
multicollinearity. If r is close to or exactly -1 or +1, one of the variables should be removed from the model if at
all possible.
Multicollinearity is also determined by the analysis of correlations between the variables and the
variance inflation factor (VIF) values (Taylor, 1990). The VIF estimates how much the variance of a regression
coefficient is inflated due to multicollinearity in the model. For the correlation coefficient, the range of values
from 0.68 to 1 is considered which was specified by Taylor in 1990 and accepted by many researchers as an
indicator of the strong correlation between the variables. As the VIF value, 4 is decided out of the values from 4,
5 and 10 which are accepted by the most researchers as indicators of upper limit that there is no multicollinearity
problem (O’Brien, 2007; Farrar et al., 1967; Wichers, 1975). Detection-tolerance or the Variance Inflation
Factor (VIF) for multicollinearity:
Tolerance= 1-Rj2; VIF = 1/tolerance
where Rj2 is the coefficient of determination of a regression of explanator j on all the other explanators. A
tolerance of less than 0.20 or 0.10 and/or a VIF of 5 or 10 and above indicates a multicollinearity.
3.6.5. Heteroscedasticity
Heteroscedasticity occurs when the variance of the errors varies across observations. If the errors are
heteroscedastic, the OLS estimator remains unbiased, but becomes inefficient. More importantly, estimates of
the standard errors are inconsistent. The estimated standard errors can be either too large or too small, in either
case resulting in incorrect inferences. Given that heteroscedasticity is a common problem in cross-sectional data
analysis, methods that correct for heteroscedasticity are important for prudent data analysis. The assumption of
homoscedasticity (meaning “same variance”) is central to linear regression models. Heteroscedasticity is
present when the size of the error term differs across values of an independent variable. Tests for
Heteroskedastic disturbances in a linear regression model is developed using the framework of the Lagrangian
multiplier test of Aitchison and Silvey (1960). It tests for the effect of the first order conditions for a maximum
likelihood of imposing the hypotheses.n statistics, maximum likelihood estimation (MLE) is a method of
estimating the parameters of a statistical model given observations, by finding the parameter values that
maximize the likelihood of making the observations given the parameters
3.7. Operationalization of Study Variables
The study variables wereoperationalized as per the previous studies as indicated below. The CG scores will be
calculated using multifactor indexes such as those used in prior studies of Bhagat et al. 2008; Bebchuk et al.
2009; Daineset al. (2010). The index will comprise eight sections. High scores for the index denote quality CG
and vice versa.
Table 3.1: Operationalization of Study Variables Variable category
& name Indicator Operational definition
Measurement Nature of
variable
Supporting
evidence from
literature
Dependent-
Pension fund
performance
Sharpe’s
ratio: Excess
Return to Variability
Composite measure of
performance, where:
St = the Sharpes index, Rp = the annually average return
on portfolio,
Rf = the risk free rate
∂p = the standard deviation of
the return of the portfolio
St =Rp –Rf
∂p
Ratio Sharpe (1964)
Pension fund value
Actual annual return of the fund assets, net or gross
Continuous
Independent –CG
composite index
Board
structure &
composition
Ownership and shareholding
(Outside ownership)
CG sub index 1
Continuous
Shleifer, A., R.
Vishny (1997);
Board size: number of trustees CG sub index 2
Board independence: percentage of outsiders in the board
CG sub index 3 Carvalhal da Silva (2005)
Independence of the chairman: if
outsider or if insider (CEO’s duality)
CG sub index 4 Carter et al.,
2003
Board diversity: measured by gender, nationality, age,
CG sub index 5 Masulis et al., 1999
Management
practices
Commitment to CG- code of
ethics
CG sub index 8
OECD (2005); Conyon& Peck
(1998)
Board procedures
Audit committees
Remuneration of directors
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 47 | Page
Variable category
& name Indicator Operational definition
Measurement Nature of
variable
Supporting
evidence from
literature
Transparency and disclosure
Certified annual financial
statements, audited and
unaudited.
CG sub index 9 Menon &
Schwartz, (1986);
OECD. (2005)
Shareholders’ right
Protection and equitable treatment of minority
shareholders
CG sub index 6 Carvalhal da
Silva (2005)
Established Legal and mutual rights of stakeholders
CG sub index 7 Carter et al., 2003
Intervening -
Investment strategy
Asset
allocation policy
Composite measure evaluated by
whether application of the investment strategies is
undertaken
Investment Strategy
index
Feldestein (1983)
Humpe &
Macmillan (2007) International
diversification
Market timing,
Portfolio
selection,
Restrictions on portfolio
performance
Moderating -
Macroeconomic factors
Gross
Domestic Products
(GDP)
Annual growth rate of the GDP Continuous
Humpe & Macmillan (2007)
Inflation rate
A general increase in prices of most goods and services
measured monthly
Consumer Price Index
Continuous Olweny &
Omondi (2011)
Interest rate
The price paid by individual or
business to borrow money measured daily
% of the shilling
borrowed
Continuous
Feldestein (1983)
Exchange rate
The rate at which one currency
will be traded for another, measured daily
Price Continuous Kane & Marcus,
2008
3.8. Data Analysis
The unit of analysis was individual pension funds. Data wasanalysed in two stages. First there was descriptive
analysis that entailed computations of frequency distributions, mean scores, standard deviations and coefficient
of variation of the fund assets value, ROA, ROE and the volatility of gross real return of the pension funds.
Secondly, the analysis involved testing for relationships between and among variables to establish their nature
and magnitude. This nvolved multiple regression analyses, Pearson’s product moment and analysis of variance
(Baron & Kenny, 1986) for the model:
Pension Financial Performance = + 1CG + 2IS + 4 MF + e.
Where CG = Corporate Governance; IS = Investment Strategy; MF = Macroeconomic factors;e = error
term. The following are the regression models and the hypotheses to be tested.
Table 3.2: Study Hypotheses and Analytical Models
Summary of Analytical Models Objectives Hypothesis Analytical Model Interpretation
Determine the influence of
Corporate
Governance (CG) on pension performance
HA: CG practices significantly
influence the
performance of pension plans in
Kenya.
H0: n =0
HA: n≠ 0
Simple regression analysis, where Pension performance =f (CG)
Y= +nXn + e
Where Y= Mean score of the Sharpe’s ratio
=Intercept/constant
n = regression coefficient (Beta) X= Aggregate mean score of the CG
= error term
Pearson’s product moment correlation R
Pearson’s product moment correlation coefficient (R) determination - The model
establishes that a set of independent
variables explains a proportion of the variance in a dependent variable at a
significant level (through a significance
test of R2). Range = +1 to -1 R= ≥ 0.7 indicates a strong positive
relationship.
Range = ≤ 0.3 indicates a weak relationship
Establish the
mediating effect (Me) of investment
Strategy (IS) on the
relationship between CG (X) and pension
H2: The investment
strategy does not mediate the effect of
CG practices on
performance of pension plans in
Path analysis/Stepwise regression
analysis: a statistical method of testing cause/effect relationships.
Step 1: Y= 0 + β1X1 + ε
Step 2: Me= 0 + β1X1 + ε
Step 1-3 establishes whether zero order
relationship among the variables exists. If one or more of these relations are not
significant, then mediation is not possible.
But if significant proceed to step 4. Full mediation is supported if CG is no
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 48 | Page
Objectives Hypothesis Analytical Model Interpretation
performance (Y) Kenya.
.
Step 3: Y=0 + β2Me + ε
Step 4: Y= 0 +β2Me + β1X1 + ε Where
Y= composite score for financial
performance
0=regression constant
X= composite score for CG
Me=mediating factor-composite score for IS
Pearson’s product moment correlation R
longer significant when IS/IC is controlled
Partial mediation is supported if both CG and IS/IC significantly predict pension
performance.
R2 to assess how much change in financial
performance is due to CG and IS or IC
If R is > 0.7 there is a positive relationship and below 0.5 there is a weak relationship.
Asses the
moderating effect of macroeconomic
factors on the
relationship between CG performance of
pension funds
HA: The influence of
CG on performance of pension funds is
significantly
moderated by macroeconomic
factors.
H0: n =0
HA: n≠ 0
Regression analysis
Y= 0 + β1X1 + ε
Y = 0 + β1X1 + β2 X2
Y = 0 + β1X1 + β2 X2 + ……..+ β1-pX1-
p.Z1-p + εi Where: Y1= Sharpe’s ratio
0 = regression coefficient and intercept
1-p = Regression coefficient s or change induced in Y by each independent
variable X
X1-p=independent variable Z1-p = moderator if the relationship
between X and Y is a function of the
level of Z
The coefficient (1-p) of the moderating
and independent variables indicate the
magnitude of the respective relationship between that variable and the first
dependent variable. Pearson’s product moment correlation R
H0: 1-p =0
HA: 1-p≠ 0
To conduct test a t test to determine
individual significance of the relationship To conduct an F test (AOV test) to assess
overall robustness and significance of the
simple regression model.
-Reject H0 if p value ≤ , otherwise fail
to reject H0 if p-value is >
Pearson’s product moment correlation coeffiInt (r)
The model establishes that a set of
independent variables explains a proportion of the variance in a dependent
variable at a significant level (through a
significance test of R2). Range = +1 to -1
R= ≥ 0.7 indicates a strong positive
relationship. Range = ≤ 0.3 indicates a weak
relationship
To determine whether the joint
effect of CG, IS, &
macroeconomic factors on pension
performance is
greater than the individual effect of
CG on pension
performance in Kenya.
The joint effect of CG, IS and
macroeconomic
factors is greater than the individual effect
of CG on pension
performance in Kenya significantly
affects the
performance of pension funds in
Kenya.
Y = 0 + β1X1 + β1-pX1-p.Z1-p + βn Men + εi
Where: Y= Sharpe’s ratio
0 = regression coefficient and intercept β1 = Regression coefficient or change
induced in Y by each independent
variable X X1=independent variable
X1-p=independent variable
Z1-p = moderator if the relationship between X and Y is a function of the
level of Z
Men= mediating variable if the relationship between X and Y is a
function of the level of Xnjn
εi = error term Pearson’s product moment correlation R
H0: 1 =2….=n=0 There is no linear relationship between Y
and the set of independent variables
HA: At least one of n ≠ 0 (There is a
linear relationship between Y and the set
of independent variables)
To conduct a t test to determine individual significance of each parameter
To conduct an F test (AOV test) to assess overall robustness and significance of the
multiple regression model.
Reject H0 if p value ≤ , otherwise fail to
reject H0 if p-value is >
If r > 0.7 with a positive and p<0.05 it
indicates CG has a positive and significant effect on pension performance.
Source: Author (2019)
IV. Research Findings 4.1. Introduction
The chapter presents study findings from the data analysis done to determine the relationship between
seven predictive factors namely corporate governance, investment strategy, macroeconomic factors (GDP
growth rate, Average interest rate and Inflation rate), the NSE 20 share index, Exchange rate and the financial
performance of pension funds in Kenya. Four hypotheses were tested: i) CG has a significant relationship with
the financial performance of pension plans; ii) Investment strategy has a significant intervening effect on the
relationship between governance and financial performance of pension plans; iii) Macroeconomic variables
(GDP growth rate, Inflation rate, Interest rate and Exchange rate have a significant moderating effect on the
relationship between governance and financial performance of pension plans; and iv) the joint effect of CG, IS,
& macroeconomic factors on pension performance is greater than the individual effect of CG on pension
performance in Kenya.Theresearch period covered the years 1997-2018 and data was obtained from industry.
4.2. Test of assumptions
The regression was done with no violation of the assumption of normality, linearity, multicollinearity,
heteroscedasticity, homoscedasticity, outliers and independence of residuals.
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 49 | Page
Table 1: Correlational for the main variables (Independent variables)
1 2 3 4 5 6 7 8 9 10
1. GDP growth rate ***
2. Inflation rate (0.2802) *** 3. NSE 20 share
index 0.4944*
0.388
9 ***
4. Maketcapitalization
0.5426**
0.2820
0.8592*** ***
5. Exchange rate Ksh
US/KS 0.4549*
(0.269
1) 0.1511 0.1767 *** 6. Equity Market
Index
0.5749*
*
0.285
6
0.9030*
**
0.6824*
** 0.2264 ***
7. Exchange rate 0.4001
(0.2489) 0.1011 0.1452
0.9785***
0.1514 ***
8. Average interest
rate
(0.5529)
**
(0.187
4)
(0.4261
)*
(0.5467)
**
(0.5538)*
*
(0.371
1) (0.5173)* ***
9. CG INDEX (0.3819)
0.368
7
(0.2224
) (0.1668)
(0.7810)
***
(0.345
9)
(0.7689)*
**
0.21
85 ***
10. IS INDEX (0.3666)
0.354
7
(0.2278
) (0.1588)
(0.7963)
***
(0.354
1)
(0.7809)*
**
0.20
83
0.9920*
**
**
*
NOTE: N=22. *P<.05, **P<.01, ***P<.001, two tailed.
Correlation coefficient indicates the strength of the relationship between the variables. Pearson’s
coefficient (r) >0.8 or > 0.9 indicate multicollinearity. The findings show that the NSE 20 share index is highly
correlated to Maket capitalisation (r=.859) and Equity Market Index (r=.903). Thus, the Maket capitalisation and
Equity Market Index were excluded in the regression analysis. Critical value calculator: The corresponding
critical correlation values rc for a significance level of α=.05; α=.01 and α=.001, for a two-tailed test are:
.05=.423; .01=.537: and .001=.652. The null hypothesis is rejected if ∣r∣>rc. Correlations greater that .433. are
statistically significant (p<.05).
Table 2: Macroeconomic variables, CG index and IS index Year GDP
growth rate
Inflation
rate
Average
interest
rate
NSE 20
share index
Exchange
rate Ksh to
US
Respondent CG INDEX IS INDEX
1997 0.48% 12.10% 22.63% 3,345 55.40 001 5.1034 5.3462
1998 3.29% 5.61% 23.23% 2,954 61.00 002 5.4483 5.4615
1999 2.31% 4.98% 16.76% 2,348 69.77 003 4.6552 4.7308 2000 0.60% 7.77% 16.59% 2,090 74.10 004 5.0345 4.9404
2001 3.78% 5.82% 14.70% 1,439 77.00 005 5.1724 5.4231
2002 0.55% 2.16% 13.69% 1,512 78.67 006 5.5000 5.8846 2003 2.90% 5.98% 11.49% 3,111 77.60 007 4.1724 4.7692
2004 5.13% 8.38% 8.44% 3,037 76.93 008 3.8276 4.6154
2005 5.90% 7.82% 9.25% 3,952 77.47 009 5.4138 5.4615 2006 6.47% 6.04% 9.45% 5,733 71.87 010 4.9655 5.2692
2007 6.85% 4.27% 9.55% 4,795 66.60 011 4.5517 5.2308
2008 0.23% 15.10% 9.75% 4,731 69.97 012 5.3103 5.0556 2009 3.31% 10.55% 10.19% 3,429 77.40 013 5.7931 5.3077
2010 8.40% 4.31% 10.00% 4,411 79.13 014 5.0714 4.9231 2011 6.10% 14.02% 10.44% 3,184 85.57 015 5.3448 5.2692
2012 4.56% 9.38% 13.70% 4,197 88.83 016 4.9655 5.2692
2013 5.88% 5.72% 12.02% 4,880 85.27 017 - - 2014 5.36% 6.88% 11.48% 5,115 86.68 018 - -
2015 5.72% 6.58% 11.33% 3,941 91.32 019 - -
2016 5.88% 6.32% 12.03% 3,173 102.08 020 - - 2017 4.86% 7.99% 11.15% 3,717 102.45 021 - -
2018 6.00% 4.69% 11.00% 2,768 102.37 022 - -
Source: KNBS 2018
Table 3: Descriptive statistics of macroeconomic variables, corporate governance and investment strategy
indices
Variable Mean
Standar
d Error Median
Standar
d
Deviati
on
Sample
Varianc
e Range
Minimu
m
Maxim
um Obs
1. GDP growth
rate 0.0448 0.0048 0.0513 0.0221 0.0005 0.0817 0.0023 0.0840 21
2. Inflation rate 0.0716 0.0068 0.0632 0.0310 0.0010 0.1294 0.0216 0.1510 21
3. NSE 20 share
index 3,548 255 3,429 1,169
1,366,50
9 4,294 1,439 5,733 21
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 50 | Page
4. Exchange rate
Ksh to US 81.05 2.52 77.60 11.54 133.25 41.45 61.00 102.45 21
5. Average
interest rate 0.1220 0.0074 0.1133 0.0341 0.0012 0.1479 0.0844 0.2323 21
6. CG INDEX 3.5822 0.5156 4.9655 2.3630 5.5836 5.7931 0 5.7931 21
7. IS INDEX 3.6958 0.5263 4.9404 2.4116 5.8159 5.8846 0 5.8846 21
Figure 1: Trends in macroeconomic variables
Figure 2: Trend in GDP growth rate , Inflation rate and Average interest rate
Source: KNBS 2018
0%
5%
10%
15%
20%
25%
PER
CEN
T
YEARS
Trend In GDP Growth Rate, Inflation Rate & Average Interest Rate
GDP growth rate Inflation rate Average interest rate
0%
5%
10%
15%
20%
25%
30%
35%
40%
PER
CEN
T
YEARS
Trend In GDP Growth Rate, Inflation Rate & Average Interest Rate
GDP growth rate Inflation rate Average interest rate
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 51 | Page
Figure 3: Trends in NSE 20 share index
Source KNBS 2018
Figure 4: Trends in Exchange rate (KS to US$)
Source KNBS 2018
-
1,000
2,000
3,000
4,000
5,000
6,000
7,000
NSE
20
SH
AR
E IN
DEX
YEARS
Trend in NSE 20 share index
-
20
40
60
80
100
120
KSH
TO
US
DO
LLA
R
YEARS
Trend in Exchange rate Ksh to US
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 52 | Page
Figure 5: Exchange rate (Ksh to US $)
Source KNBS 2018
Table 4: Returns of pension funds
Source: Pension schemes, 2018
-
20
40
60
80
100
120
KSH
TO
US
$
YEARS
Exchange rate Ksh to US
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 53 | Page
Table 5: Descriptive Statistics of the pension funds
Statistic
NSSF
Fund
Balance
in KS
'000'
TELPO
STA
PENSI
ON
SCHE
ME in
KS
'000'
LAP
TRU
ST
inKs
h
'000'
CPF
INDIVI
DUAL
PENSI
ON in
KS '000'
CPF
INDIVI
DUAL
PENSI
ON in
KS
'000'
Amana
Umbrell
a
Pension
in KS
'000'
KENY
APOW
ER
PENSI
ON
FUND
(Umbr
ella) in
KS
'000'
KENY
APOW
ER
PENSI
ON
FUND
in KS
'000'
THE
JUBIL
EE
INSUR
ANCE
UMB
in KS
'000'
ICEA
LION
LIFE
ASSU
RAN
CE in
KS
'000'
The
Her
itag
e
Ins
ura
nce
Grp
-
Fun
d
valu
ein
KS
'000
'
The
Her
itag
e
Life
Ins
ura
nce
-
Fun
d
valu
e in
KS
'000
'
The
Heri
tage
Insu
ranc
e
Grp
-
Tota
l
asse
t
in
KS
'000
'
The
Her
itag
e
Life
Ins
ura
nce
-
Tot
al
in
KS
'000
'
Mean
84,436,7
13
5,854,9
94
9,53
0
7,194,10
0
3,666,3
52
4,047,34
2 7,428 3,662
410,89
0
15,72
7
1,82
5
1,42
7
6,01
2
4,02
1
Standar
d Error
14,768,6
05
1,451,3
17
2,25
6
2,569,41
0
1,659,5
99
2,793,40
6 1,734 1,033
286,43
5 4,872 270 213 892 627
Median
81,729,3
78 -
5,30
0 - - - 4,083 603 - -
2,08
0
1,74
5
6,75
5
4,50
0
Standar
d
Deviati
on
69,270,8
95
6,807,2
79
10,5
84
11,774,5
17
7,784,2
10
13,102,2
36 8,132 4,847
1,343,5
00
22,85
4
1,26
7 998
4,18
6
2,93
9
Sample
Varianc
e
4,798,45
6,940,86
5,780
46,339,
044,780
,061
112,
014,
621
138,639,
255,613,
953
60,593,
932,574
,827
171,668,
582,349,
427
66,121
,794
23,494
,013
1,804,9
91,124,
626
522,2
84,27
8
1,60
4,77
6
996,
474
17,5
19,7
56
8,63
6,73
8
Kurtosis (0.790) (1.773)
(1.46
4) (0.766) 2.845 8.103 (1.728) (0.148) 9.133
(0.32
7)
(1.0
65)
(0.7
80)
(1.2
66)
(0.8
87)
Skewne
ss 0.336 0.406
0.50
5 1.096 2.078 3.061 0.352 1.096 3.170 1.052
(0.4
25)
(0.3
09)
(0.4
88)
(0.1
21)
Range
224,024,
782
15,969,
424
27,7
25
29,944,9
44
22,651,
603
45,303,2
06 19,478 14,200
5,137,0
56
68,49
1
3,66
2
3,10
2
11,6
33
9,20
0
Minimu
m - - - - - - - - - - - - - -
Maximu
m
224,024,
782
15,969,
424
27,7
25
29,944,9
44
22,651,
603
45,303,2
06 19,478 14,200
5,137,0
56
68,49
1
3,66
2
3,10
2
11,6
33
9,20
0
Sum
1,857,60
7,680
128,809
,870
209,
650
151,076,
106
80,659,
746
89,041,5
34
163,40
6 80,567
9,039,5
87
345,9
95
40,1
40
31,3
89
132,
258
88,4
57
Observ. 22 22 22 21 22 22 22 22 22 22 22 22 22 22
Statistic
Libert
y Life
Assur
ance
Kenya
Ltd F
in KS
'000'
Total
assets
Liber
ty
Life
Assur
ance
in KS
'000'
Fund
value
Libe
rty
Hold
ings
in
Ksh
"000
"
Net
asset
s
Libert
y
Holdi
ngs in
Ksh
"000"
Total
assets
Brita
m Life
Assur
ance
in KS
'000'
BRITAM
Individua
lPension
in KS
'000'
Britam
Umbrella
Pension in
KS '000'
Britam
holdin
gs
consoli
dated
in KS
'000'
CfC
life
Insur
ance
Holdi
ngS in
'000'
CfC
Stanbic
Holdin
gs
Limite
d in
Ksh
'000'
CIC
Life
Assur
ance
in
Ksh
'000'
CIC
Insura
nce
group
limit in
KS
'000'(T
otal
assets)
Jubele
e
Holdin
gs in
KS
'000'
Mean 7,115 6,453
2,40
4 13,457 7,237 7,142 247,134 28,108 8,938 98,147 2,912 9,117 35,597
Standard
Error 2,288 2,078 650 3,395 3,098 4,509 171,678 7,569 3,468 20,084 819 2,445 7,659
Median - - - - - - - 13,652 - 77,196 675 2,734 19,050
Standard
Deviation 10,733 9,748
3,04
9 15,924 14,532 21,149 786,729 35,504 13,433 94,204 3,841 11,469 35,925
Sample
Variance
115,19
6,838
95,01
4,754
9,29
5,64
4
253,56
0,547
211,18
4,259
447,296,4
98
618,942,411
,293
1,260,5
02,058
180,45
2,778
8,874,3
64,174
14,75
1,568
131,54
3,386
1,290,6
03,530
Kurtosis
(1.294
)
(1.27
5)
(1.45
0)
(1.764
) 1.940 11.350 8.354 (0.299)
(0.984
) (1.076) 0.201 (0.550) (0.260)
Skewness 0.882 0.889
0.63
8 0.452 1.808 3.324 3.060 1.079 0.964 0.486 1.166 1.003 1.049
Range 24,446
22,08
1
7,61
9 37,339 45,628 88,724 2,906,637
101,50
0 33,194
290,57
0
12,18
5 32,976
109,78
0
Minimum - - - - - - - - - - - - 4,390
Maximum 24,446
22,08
1
7,61
9 37,339 45,628 88,724 2,906,637
101,50
0 33,194
290,57
0
12,18
5 32,976
114,17
0
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 54 | Page
Sum
156,53
7
141,9
71
52,8
79
296,06
3
159,21
1 157,133 5,189,812
618,38
5
134,07
1
2,159,2
23
64,06
4
200,56
5
783,13
8
Observ. 22 22 22 22 22 22 21 22 15 22 22 22 22
The findings show that distribution is “normal,” almost all (96%) of your observations fall within +/- 2 standard
deviations from the mean.
4.3. Hypotheses testing
4.3.1. The effect of Corporate Governance (CG) on financial performance of pension plans
The first hypothesis of the study was to test and establish the effect of Corporate Governance (CG) on the
financial performance of pension plans in Kenya. Regression analysis was used and the findings presented in
tables 4-8 for the various pension funds.
4.3.4.1. Regression statistics on the effect of CG on performance of the NSSF
Table 6: Summary Output of the effect of CG index on NSSF performance Regression Statistics
Multiple R 0.8211 R Square 0.6742
Adjusted R Square 0.6579
Standard Error 42,668,332 Observations 22
ANOVA
Df SS MS F Significance F
Regression 1 75,337,914,288,764,500 75,337,914,288,764,500 41.38 0.0000028
Residual 20 36,411,730,496,989,300 1,820,586,524,849,470
Total 21 111,749,644,785,754,000 Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept 181,107,884 17,201,655 10.5285 0.0000000 145,225,860 216,989,908
CG index -25,720,614 3,998,344 -6.4328 0.0000028 -34,061,014 -17,380,214
i. Predictor: CG index
ii. Dependent variable: NSSF fund value
A multiple regression was carried out to investigate effect of Corporate Governance (CG) on financial
performance of pension plans. The results of the regression indicated that the model explained 67.42% of the
variation in the independent variable, the NSSF fund value. The study findings also showed that the model was
a significant predictor of pension funding, F (1,20) = 41.38, p < .001. CG index contributed significantly to the
model (t = -6.43, p<.001). The final predictive model was: NSSF funding level = 181,107,884 –
25,720,614*CG index.
4.3.4.2. Regression statistics on the effect of CG on performance of Teleposta, LAPTRUST, Heritage
Insurance and CfC Stanbic pension schemes
Table 7: Summary output of the Effect of CG on Teleposta pension fund performance Regression Statistics
Multiple R 0.503 R Square 0.253
Adjusted R Square 0.216
Standard Error 5,549,487 Observations 22
ANOVA
Df SS MS F Significance F
Regression 1 208,665,180,097,711 208,665,180,097,711 6.78 0.0170 Residual 20 615,936,076,733,697 30,796,803,836,685
Total 21 824,601,256,831,409
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept 9,472,702 2,237,265 4.23405 0.00041 4,805,849 14,139,554
CG index (1,353,629) 520,029 (2.60299) 0.01702 (2,438,390) (268,868)
Table 8: Summary output of the Effect of CG on performance of LAPTRUST pension fund Regression Statistics
Multiple R 0.805
R Square 0.648 Adjusted R Square 0.630
Standard Error 6,437
Observations 22 ANOVA
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 55 | Page
Df SS MS F Significance F
Regression 1 1,523,530,772 1,523,530,772 37 0.0000063
Residual 20 828,776,274 41,438,814
Total 21 2,352,307,045
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept 22,885 2,595 8.818217981 0.000000025 17,471 28,298
CG index (3,658) 603 (6.063480123) 0.000006303 (4,916) (2,399)
Table 9: Summary output of the Effect of CG on the performance of Heritage Insurance fund
Regression Statistics
Multiple R 0.6701
R Square 0.4490
Adjusted R
Square 0.4214
Standard Error 759
Observations 22
ANOVA
df SS MS F
Significance
F
Regression 1 9,395,184 9,395,184 16 0.0006455
Residual 20 11,530,767 576,538
Total 21 20925951.86
Coefficient
s
Standard
Error t Stat P-value Lower 95%
Upper
95%
Intercept 2,476 306 8.08709798
0.0000001
0 1,837 3,114
CG index (287) 71
(4.03681218
)
0.0006455
0 (436) (139)
Table 10: Summary output of the Effect of CG on the performance of the CfC Stanbic Holdings pension
fund Regression Statistics
Multiple R 0.7751 R Square 0.6008
Adjusted R Square 0.5808
Standard Error 60,991 Observations 22
ANOVA
df SS MS F Significance F
Regression 1 111,963,792,226 111,963,792,226 30 0.0000227 Residual 20 74,397,855,437 3,719,892,772
Total 21 186,361,647,664
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept 212,637 24,588 8.647860 0.000000034 161,346 263,927
CG index (31,355) 5,715 (5.486225) 0.000022717 (43,277) (19,434)
The study established that the R2 for the overall models of the other funds namely Teleposta,
LAPTRUST, Heritage Insurance and CfC Stanbic pensionwere 25.3%; 64.8%; 44.9%; and 60.08% with an
adjusted R2 of 21.6%; 63%; 42.14%; and 58.08% respectively as indicated in tables 6-9. These results imply that
the models explained 25-65% of the variation in the independent variable, the pension fund value depending on
the fund being tested.Thus 25-65% of the variance in the dependent variable (NSSF value) is explained by the
independent variables (CG index) in the model. In addition, the study established that the models were
significant predictors of pension funding: Teleposta- F (1,20) = 6.78, p < .05; LAPTRUST- F (1,20) = 37, p <
.001; Heritage Insurance- F (1,20) = 16, p < .001 and CfC Stanbic pension- F (1,20) = 30, p < .001. The
coefficient statistics that tells us the extent to which the individual predictor variables contribute to the
model,indicated that CG index contributed significantly to the four models of the various pension funds:
Teleposta t = -2.6, p <.05; LAPTRUST t = -6.06, p <.001; Heritage Insurance t = -4.04, p <.001; and CfC
Stanbic pension t = -5.49, p <.001. These findings show that CG is a useful predictor of pension funding level as
it has a significant impact on the funding levels. This implies that the tested hypotheses is accepted: H1:
Corporate governance (CG) has a significant relationship with the financial performance of pension plans.
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 56 | Page
4.3.2. The effect of Investment Strategy (IS) (mediator) on the relationship between corporate
governance and financial performance of pension plans
Mediation analysis using multiple regression was carried out to investigate the effect of Investment
Strategy (IS) (mediator) on the relationship between corporate governance and financial performance of pension
plans. The regression process involved testing mediational hypotheses through the 4 steps as outlined by Baron
and Kenny (1986), Judd and Kenny (1981), and James and Brett (1984). The study findings are indicated in the
tables below.
4.3.4.1. The effect of CG on NSSF funding
Table 11: Step 1: Summary output of the effect of CG on NSSF funding Regression Statistics
Multiple R 0.8211 R Square 0.6742
Adjusted R Square 0.6579
Standard Error 42,668,332 Observations 22
ANOVA
Df SS MS F Significance F
Regression 1 75,337,914,288,764,500 75,337,914,288,764,500 41.38 0.0000028
Residual 20 36,411,730,496,989,300 1,820,586,524,849,470
Total 21 111,749,644,785,754,000
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept 181,107,884 17,201,655 10.5285148 0.0000000 145,225,860 216,989,908
CG index (25,720,614) 3,998,344 (6.4328161) 0.0000028 (34,061,014) (17,380,214)
The results of the regression indicate that the model explained 67.42% of the variation in the NSSF fund value.
In addition, the model was a significant predictor of NSSF funding level as shown by the F (1,20) = 41.38, p <
.001. CG index contributed significantly to the model (t = -6.43, p<.001). Corporate governance is thus
individually useful in the prediction of the NSSF fund level. The final predictive model was:
NSSF funding level = 181,107,884 – 25,720,614*CG index.
4.3.4.2. The effect of corporate governance (CG index) on investment strategy (IS index) (mediating
factor)
Table 12: Step 2: Summary output of the effect of CG on IS (mediating factor) Regression Statistics
Multiple R 0.99197
R Square 0.98400
Adjusted R Square 0.98320 Standard Error 0.30840
Observations 22
ANOVA
Df SS MS F Significance F
Regression 1 117 117 1,230 0.0000
Residual 20 2 0
Total 21 119 Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept 0.06951 0.12433 0.55907 0.58232 (0.18984) 0.32886
CG index 1.01367 0.02890 35.07541 0.00000 0.95339 1.07396
The study findings show that the R2 for the overall model was 98.4% with an adjusted R
2 of 98.3%. A high size
effect is reported of the model. The results imply that the model explained 98.4% of the variation in the
dependent variable, IS index. In addition, the model was a significant predictor of the IS index as shown by the
F (1,20) = 1,230, p < .001. The coefficient statistics show that CG index contributed significantly to the model (t
= 35.08, p <.001). Corporate governance is thus individually useful in the prediction of the IS index. The final
predictive model was: IS index = 0.0695+ 1.0137*CG index.
4.3.4.3. The effect of investment strategy (IS) (mediating factor) on NSSF funding level
Table 13: Step 3: Summary output of the effect of IS (mediating factor) on NSSF level Regression Statistics
Multiple R 0.84
R Square 0.70 Adjusted R Square 0.69
Standard Error 40,662,844
Observations 22 ANOVA
Df SS MS F Significance F
Regression 1 78,680,306,957,670,700 78,680,306,957,670,700 48 0.0000011
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 57 | Page
Residual 20 33,069,337,828,083,200 1,653,466,891,404,160
Total 21 111,749,644,785,754,000 Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept 184,185,824 16,518,489 11.1502826 0.0000000 149,728,859 218,642,788
IS index (25,722,232) 3,728,837 (6.8981919) 0.0000011 (33,500,449) (17,944,014)
The regression statistics showed that the R2 for the overall model was 70% with an adjusted R
2 of 69%. A high
size effect is reported of the model. IS index thus explains 70% of the variation in the dependent variable, NSSF
funding level. The study findings also show that the model was a significant predictor of the NSSF funding level
as indicated by the F (1,20) = 48, p < .001. The model therefore has explanatory power. The coefficient statistics
show that IS index contributed significantly to the model (t = = -6.9, p <.001). Investment Strategy is thus
individually useful in the prediction of the NSSF funding level. The final predictive model was: NSSF funding
level = 184,185,824 -25,722,232*IS index.
The study findings show that Step 1-3 establishes that there is a linear relationship among the variables NSSF
funding level, CG index and IS index. All the relations are significant indicating that mediation is possible
allowing proceed to step 4.
4.3.4.4. The effect of CG index and IS index on NSSF funding level
Table 14: Step 4: Summary output of the effect of CG and IS indices on the NSSF level
Regression Statistics
Multiple R 0.84
R Square 0.71
Adjusted R Square 0.68
Standard Error
41,154,9
14
Observations 22
ANOVA
df SS MS F
Significan
ce F
Regression 2
79,568,832,814,39
3,800
39,784,416,407,1
96,900 23 0.0000073
Residual 19
32,180,811,971,36
0,100
1,693,726,945,86
1,060
Total 21
111,749,644,785,7
54,000
Coefficie
nts Standard Error t Stat
P-
value
Lower
95%
Upper
95%
Intercept
184,386,
094 16,720,669 11.027435
0.0000
00
149,389,3
31
219,382,
858
Corporate
Governance Index
22,085,0
80 30,491,981 0.724291
0.4777
12
(41,735,37
0)
85,905,5
31
Investment Strategy
Index
(47,160,9
30) 29,839,169 (1.580504)
0.1304
96
(109,615,0
30)
15,293,1
69
Regression statisticsshow that the R2 for the overall model was 71% with an adjusted R
2 of 68%. A high size
effect is reported of the model. The two independent variables CG index and IS index explain 71% of the
variation in the dependent variable, NSSF funding level. The ANOVA analysis showed that the model was a
significant predictor of the NSSF funding level as indicated by the F (2,19) = 23, p < .001. The model therefore
has explanatory power. The coefficient statistics however, show that CG index and IS index did not contribute
significantly to the model (CG index t = 0.72, p = 0.478; IS index t = -1.58, p = 0.130). In both cases the p value
> .05. The two factors are therefore not individually significant in the prediction of the NSSF fund level. The
final predictive model was:
NSSF fund level = 184,386,094+ 22,085,080*CG index - 47,160,930*IS index.
4.3.4.5. The effect of CG and IS indices on the CfC funding level
Table 15: Step 4: Summary output of the effect of CG index & IS index on the funding level of the CfC
pension fund
Regression Statistics
Multiple R 0.85
R Square 0.73
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 58 | Page
Adjusted R
Square 0.70
Standard Error 51,752
Observations 22
ANOVA
df SS MS F
Significance
F
Regression 2
135,475,201,49
8
67,737,600,74
9 25 0.0000044
Residual 19 50,886,446,166 2,678,234,009
Total 21
186,361,647,66
4
Coefficient
s Standard Error t Stat
P-
value Lower 95%
Upper
95%
Intercept 220,365 21,026 10.4806 0.0000 176,357 264,373
CG index 81,339 38,343 2.1213 0.0473 1,086 161,592
IS index (111,174) 37,522 (2.9629) 0.0080 (189,709) (32,639)
Regression statisticsshow that the R2 for the overall model was 73% with an adjusted R
2 of 70%. A high size
effect is reported of the model. The findings imply that the two independent variables CG index and IS index
explain 73% of the variation in the dependent variable, CfC funding level. The ANOVA analysis showed that
the model was a significant predictor of the CfC funding level as indicated by the F (2,19) = 25, p<.001. The
model therefore has explanatory power. In contrast to the NSSF, the coefficient statistics show that CG index
and IS index contributed significantly to the model (CG index t = 2.12, p < .05; IS index t = -2.96, p < .05). The
two factors are therefore individually useful in the prediction of the CfC fund level. The final predictive model
was:
CfC fund level = 220,365 + 81,339*CG index - 111,174*IS index.
The above study findings show that investment strategy has a significant mediating effect on the performance of
pension funds though its individual contribution in the model varies. This implies that the tested hypotheses is
accepted: H2: Investment strategy has a significant intervening/mediating effect on the relationship between
governance and financial performance of pension plans.
4.3.3. The effect of macroeconomic variables on the relationship between governance and financial
performance of pension plans.
The third hypothesis of the study was to test and establish the effect of macroeconomic variables on the
relationship between governance and financial performance of pension plansRegression analysis was used and
the findings presented in tables 14-23 for the various pension funds.
Table 16: Summary Output of the effect of macroeconomic variables on the relationship between
governance and performance of the NSSF fund Regression Statistics
Multiple R 0.969
R Square 0.939 Adjusted R Square 0.915
Standard Error 21,306,237
Observations 22 ANOVA
Df SS MS F
Significance
F
Regression 6 104,940,309,049,676,0
00 17,490,051,508,279,3
00 39 0.000000028
Residual 15 6,809,335,736,077,780 453,955,715,738,519
Total 21
111,749,644,785,754,000
Coefficients Standard Error t Stat
P-
value Lower 95% Upper 95%
Intercept
(229,107,863
) 100,116,787 (2.2884) 0.0370 (442,501,744) (15,713,982)
GDP growth rate 401,766,548 313,938,359 1.2798 0.2201 (267,377,226) 1,070,910,32
1
Inflation rate 391,174,224 178,487,609 2.1916 0.0446 10,736,892 771,611,557
Average interest rate (4,403,605) 173,268,068 (0.0254) 0.9801 (373,715,751) 364,908,540 NSE 20 share index 16,400 6,580 2.4923 0.0249 2,374 30,425
Exchange rate Ksh to 3,164,111 810,833 3.9023 0.0014 1,435,861 4,892,361
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 59 | Page
US
CG INDEX (10,983,574) 3,795,992 (2.8935) 0.0111 (19,074,539) (2,892,610)
Table 17: Summary output of the effect of macroeconomic variables on the relationship between
governance and financial performance of theTeleposta fund Regression Statistics
Multiple R 0.9395
R Square 0.8826
Adjusted R Square 0.8356 Standard Error 2,759,923.3637
Observations 22.0000
ANOVA
df SS MS F Significance F
Regression 6 858,862,285,778,900 143,143,714,296,483 18.7922 0.00000342
Residual 15 114,257,654,602,386 7,617,176,973,492 Total 21 973,119,940,381,286
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (46,080,606) 12,968,722 (3.55321) 0.00289 (73,722,783) (18,438,430)
GDP growth rate 65,033,785 40,666,300 1.59921 0.13062 (21,644,382) 151,711,952 Inflation rate 36,084,049 23,120,560 1.56069 0.13944 (13,196,258) 85,364,356
Average interest rate 60,937,498 22,444,442 2.71504 0.01597 13,098,302 108,776,694
NSE 20 share index 1,337 852 1.56810 0.13771 (480) 3,153 Exchange rate Ksh to US 446,550 105,032 4.25156 0.00070 222,680 670,421
CG INDEX (452,452) 491,717 (0.92015) 0.37206 (1,500,523) 595,618
Table 18: Summary output effect of macroeconomic factors on the relationship between governance and
financial performance of the LAPTRUST pension fund Regression Statistics
Multiple R 0.9594 R Square 0.9205
Adjusted R Square 0.8887
Standard Error 3,531.26 Observations 22.00
ANOVA Df SS MS F Significance F
Regression 6 2,165,259,829 360,876,638 28.9400 0.00000020
Residual 15 187,047,217 12,469,814
Total 21 2,352,307,045 Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (70,995) 16,593 (4.27856) 0.00066 (106,363) (35,627)
GDP growth rate 18,999 52,032 0.3651 0.72011 (91,904) 129,902 Inflation rate 67,856 29,582 2.2938 0.03666 4,803 130,909
Average interest rate 79,860 28,717 2.7809 0.01399 18,651 141,070
NSE 20 share index 2.8451 1.0906 2.6088 0.01975 0.5206 5.1697 Exchange rate Ksh to US 723 134 5.3825 0.00008 437 1,010
CG INDEX (898) 629 (1.4275) 0.17393 (2,239) 443
Table 19: Summary outputeffect of macroeconomic factors on the relationship between governance
andfinancial performance of the Kenya Power Company DB pension fund Regression Statistics
Multiple R 0.9651 R Square 0.9315
Adjusted R Square 0.9041
Standard Error 1,501 Observations 22
ANOVA df SS MS F Significance F
Regression 6 459,581,037 76,596,839 34.00 0.00000007 Residual 15 33,793,233 2,252,882
Total 21 493,374,270
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (22,922) 7,053 (3.24999) 0.00538 (37,955) (7,889)
GDP growth rate (6,621) 22,116 (0.29936) 0.76877 (53,760) 40,519
Inflation rate 18,888 12,574 1.50214 0.15382 (7,913) 45,688 Average interest rate 29,090 12,206 2.38325 0.03082 3,074 55,107
NSE 20 share index 0.8093 0.4636 1.74578 0.10129 (0.1788) 1.7973
Exchange rate Ksh to US 275 57 4.81706 0.00023 153 397 CG INDEX (838) 267 (3.13236) 0.00685 (1,408) (268)
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 60 | Page
Table 20: Summary outputeffect of macroeconomic factors on the relationship between governance
andfinancial performance of the Kenya Power Company DC pension fund Regression Statistics
Multiple R 0.9442
R Square 0.8916 Adjusted R Square 0.8482
Standard Error 3,168
Observations 22 ANOVA
df SS MS F Significance F
Regression 6 1,238,023,096 206,337,183 20.56 0.0000019 Residual 15 150,534,582 10,035,639
Total 21 1,388,557,678
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (54,988) 14,886 (3.6940) 0.0022 (86,716) (23,260)
GDP growth rate 14,443 46,678 0.3094 0.7613 (85,048) 113,934
Inflation rate 53,905 26,538 2.0312 0.0603 (2,661) 110,470 Average interest rate 60,035 25,762 2.3303 0.0342 5,124 114,946
NSE 20 share index 2.3928 0.9784 2.4457 0.0273 0.3075 4.4782
Exchange rate Ksh to US 550 121 4.5639 0.0004 293 807 CG INDEX (608) 564 (1.0767) 0.2986 (1,811) 595
Table 21: Summary outputeffect of macroeconomic factors on the relationship between governance
andfinancial performance of theHeritage (Grp) Regression Statistics
Multiple R 0.9704
R Square 0.9417 Adjusted R Square 0.9184
Standard Error 362
Observations 22 ANOVA
df SS MS F Significance F
Regression 6 31,735,539 5,289,256 40.38 0.000000020
Residual 15 1,964,759 130,984 Total 21 33,700,297
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (3,363) 1,701 (1.9772) 0.0667 (6,987) 262 GDP growth rate 10,267 5,333 1.9253 0.0734 (1,100) 21,633
Inflation rate 6,107 3,032 2.0142 0.0623 (356) 12,569
Average interest rate (4,866) 2,943 (1.6532) 0.1191 (11,139) 1,408 NSE 20 share index 0.3558 0.1118 3.1836 0.0062 0.1176 0.5941
Exchange rate Ksh to US 50 14 3.6612 0.0023 21 80
CG INDEX (103) 64 (1.5978) 0.1309 (240) 34
Table 22: Summary output effect of macroeconomic factors on the relationship between governance
andfinancial performance of theHeritage insurance pension fund Regression Statistics
Multiple R 0.9370
R Square 0.8779
Adjusted R Square 0.8291 Standard Error 413
Observations 22
ANOVA
df SS MS F Significance F
Regression 6 18,370,882 3,061,814 17.97 0.00000455
Residual 15 2,555,070 170,338 Total 21 20,925,952
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (1,562) 1,939 (0.8053) 0.4332 (5,695) 2,572
GDP growth rate 7,256 6,081 1.1931 0.2514 (5,706) 20,217 Inflation rate 3,604 3,457 1.0423 0.3138 (3,766) 10,973
Average interest rate (6,065) 3,356 (1.8069) 0.0909 (13,218) 1,089 NSE 20 share index 0.2259 0.1275 1.7723 0.0967 (0.0458) 0.4976
Exchange rate Ksh to US 33.83 15.71 2.15 0.0479 0.35 67.31
CG INDEX (88.43) 73.53 (1.20) 0.2478 (245.16) 68.30
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 61 | Page
Table 23: Summary outputeffect of macroeconomic factors on the relationship between governance
andfinancial performance of theLiberty pension fund Regression Statistics
Multiple R 0.9805
R Square 0.9614 Adjusted R Square 0.9459
Standard Error 2,496
Observations 22 ANOVA
df SS MS F Significance F
Regression 6 2,325,669,101 387,611,517 62.21 0.0000000010 Residual 15 93,464,508 6,230,967
Total 21 2,419,133,608
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (55,442) 11,729 (4.7267) 0.00027 (80,442) (30,441)
GDP growth rate (52,341) 36,780 (1.4231) 0.17518 (130,736) 26,054
Inflation rate (10,922) 20,911 (0.5223) 0.60908 (55,493) 33,649 Average interest rate 95,816 20,300 4.7201 0.00027 52,548 139,084
NSE 20 share index 3.2834 0.7709 4.2590 0.00069 1.6402 4.9265
Exchange rate Ksh to US 613 95 6.4496 0.00001 410 815 CG INDEX (1,944) 445 (4.3704) 0.00055 (2,892) (996)
Table 24: Summary outputeffect of macroeconomic factors on the relationship between governance
andfinancial performance of theCFC (GRP) Regression Statistics
Multiple R 0.9777
R Square 0.9559 Adjusted R Square 0.9383
Standard Error 2,850
Observations 22 ANOVA
df SS MS F Significance F
Regression 6 2,640,595,904 440,099,317 54.19 0.0000000025
Residual 15 121,815,199 8,121,013 Total 21 2,762,411,103
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (54,235) 13,391 (4.0502) 0.0010 (82,777) (25,693) GDP growth rate 40,691 41,990 0.9691 0.3479 (48,808) 130,190
Inflation rate 43,484 23,873 1.8215 0.0885 (7,400) 94,368
Average interest rate 76,351 23,175 3.2946 0.0049 26,955 125,747 NSE 20 share index 1.7419 0.8801 1.9791 0.0665 (0.1340) 3.6178
Exchange rate Ksh to US 624 108 5.7512 0.0000 393 855
CG INDEX (1,993) 508 (3.9255) 0.0013 (3,075) (911)
Table 25: Summary outputeffect of macroeconomic factors on the relationship between governance
andfinancial performance of theCFC assurance fund Regression Statistics
Multiple R 0.9625
R Square 0.9264
Adjusted R Square 0.8970 Standard Error 1,233
Observations 22
ANOVA
Df SS MS F Significance F
Regression 6 286,981,436 47,830,239 31.47 0.00000011
Residual 15 22,801,500 1,520,100 Total 21 309,782,936
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (18,671) 5,793 (3.2227) 0.0057 (31,019) (6,322)
GDP growth rate 21,617 18,167 1.1899 0.2526 (17,104) 60,338 Inflation rate 16,221 10,329 1.5705 0.1371 (5,793) 38,236
Average interest rate 26,366 10,026 2.6296 0.0189 4,995 47,737 NSE 20 share index 0.3548 0.3808 0.9317 0.3662 (0.4568) 1.1664
Exchange rate Ksh to US 214 47 4.5530 0.0004 114 314
CG INDEX (605) 220 (2.7529) 0.0148 (1,073) (137)
i. Predictors: GDP growth rate, Inflation rate, NSE 20 share index, Exchange rate, Average interest rate and
CG index.
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 62 | Page
ii. Dependent variable: Pension fund value
4.3.4.1. The effect of macroeconomic variables (moderators) on the relationship between governance and
financial performance of pension plans
A multiple regression was carried out to investigate effect of macroeconomic variables (moderators) on
the relationship between corporate governance and financial performance of pension plans. The results of the
regression indicated that the Pearson’s product moment correlation (r) ranged between 0.82-0.97 for the tested
pension funds indicating a strong positive relationship. In addition, the R2 for the overall model for the various
pension funds ranged between 86-97% with an adjusted R2 ranging between 82-95%. The models therefore
explained 86-97% of the variance of the pension fund level. Thus, the combined effect of GDP growth rate,
Inflation rate, Average interest rate, NSE 20 share index, Exchange rate and CG index explain 86-97% of the
variation in the independent variable, pension fund value.
The study also established that the model was a significant predictor of pension fund value as indicated
by the F statistics in the ANOVA Tables 14-23: NSSF -F (6,15) = 39, p values <.001;Teleposta -F (6,15) =
18.79, p values <.001; LAPTRUST -F (6,15) = 28.94, p values <.001, KPLCI Individual- F (6,15) = 34, p values
<.001; KPLCI umbrella- F (6,15) = 20.56, p values <.001; Heritage (Grp) - F (6,15) = 40.38, p values <.001;
Heritage Life insurance- F (6,15) = 17.97, p values <.001: Liberty fund- F (6,15) = 62.21, p values <.001; CfC
(Grp)- F (6,15) = 54.19, p values <.001 and CfC Life assurance fund F (6,15) = 31.47, p values <.001. The
results show strong evidence to reject the null hypotheses that the coefficient is equal to zero (no effect). The
results thus indicate that there is significant regression relationship between the dependent variable (pension
funding level) and the predictor variables (macroeconomic variables and CG index) as is indicated by a large F
values and a small significance level.
The relative importance of the independent variables is judged for by the magnitude of the t statistics.
The study findings indicate that thet test is statistically significant for the tested macroeconomic variables and
CG index because of their high t values and small significance levels. The effects of individual predictor
variables on the relationship however, varied with the pension fund as indicated by the Coefficient Tables 14-
23. Exchange rate was found to be individually useful in the prediction of all (10) the assessed pension fund
levels: t ranged between 2.15-6.44, p value<.05. Average interest rate was individually valuable in the
prediction 7 pension funds : Teleposta (t=2.72, p value<.05), LAPTRUST(t=2.78, p value < .05), KPLCI
Individual(t=2.38, p value < .05), KPLCI umbrella (t=2.33, p value < .05), Liberty funds (t=4.72, p value <
.001), CfC (Grp) (t=3.29, p value < .05), and CfC Life insurance fund(t=2.63, p value < .05). The NSE 20 share
index on the other hand was found to be individually useful in the prediction of 5 pension funds: the NSSF
(t=2.49, p value < .05), LAPTRUST(t=2.61, p value < .05), KPLC umbrella (t=2.45, p value < .05), Heritage
(GRP) (t=3.18, p value < .05), Liberty fund (t=4.26, p value < .001).
The study also established that Inflation rate was found to be individually significant in the prediction
of 2 pension funds: the LAPTRUST (t=2.29, p value < .05), and NSSF (t=2.19, p value < .05) whereas CG
index was significant in the prediction of 5 schemes: the NSSF (t=-2.89, p value < .05), KPLC individual (t=-
3.12, p value < .05), Liberty fund (t= -4.37, p value < .001), CfC ( Grp) (t= - 3.93, p value < .001), CfC fund
(t=-2.75, p value < .05). The study nonetheless established that GDP growth rate was not individually significant
in the prediction of any of the funds funding levels.
An evaluation of the p-values suggests that four independent variables are statistically significant in the
prediction of fund levels with the exception of GDP growth rate. The findings show that exchange rate is the
most significant independent variable, followed by Average interest rate and NSE 20 share index. GDP growth
rate does not reach statistical significance (p>.05) in the multiple regression model in all the tested funds. These
results imply that the influence of individual macroeconomic factors and CG index in the prediction of fund
levels varies. In general, the study establishes the acceptance of three hypotheses involving inflation rate,
interest rate and exchange rate and rejection of one involving GDP growth rate suggesting that:
a) Inflation rate has a significant moderating effect on the relationship between CG practices and financial
performance of pension plans.
b) Interest rate has a significant moderating effect on the relationship between CG practices and financial
performance of pension plans.
c) Exchange rate has a significant moderating effect on the relationship between CG practices and fiscal
performance of pension plans.
d) GDP growth rate has no significant moderating effect on the association between CG practices and
financial performance of pension plans.
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 63 | Page
4.3.4. The joint effect of CG, IS, & macroeconomic factors on the performance of pension funds in
Kenya.
The fourth hypothesis of the study was to investigate whether the joint effect of CG index, IS index, &
macroeconomic factors on pension performance is greater than the individual effect of CG on pension
performance in Kenya.Regression analysis was used and the findings presented in tables 24-36 for the various
pension funds.
Table 26:Summary output of the effect of CG index, IS index, & macroeconomic factors on the financial
performance NSSF Regression Statistics
Multiple R 0.9713
R Square 0.9434
Adjusted R Square 0.9151 Standard Error 21,256,994
Observations 22
ANOVA
Df SS MS F
Significance
F
Regression 7 105,423,607,676,679,0
00 15,060,515,382,382,7
00 33 0.00000011
Residual 14 6,326,037,109,075,290 451,859,793,505,378
Total 21
111,749,644,785,754,000
Coefficients Standard Error t Stat
P-
value Lower 95% Upper 95%
Intercept (185,990,325
) 108,237,139 (1.7184) 0.1078 (418,135,899) 46,155,249
GDP growth rate 427,424,473 314,193,821 1.3604 0.1952 (246,454,252)
1,101,303,19
7 Inflation rate 338,823,452 185,129,857 1.8302 0.0886 (58,240,601) 735,887,505
Average interest rate (57,278,894) 180,269,632 (0.3177) 0.7554 (443,918,801) 329,361,014
NSE 20 share index 14,990 6,705 2.2356 0.0422 609 29,371 Exchange rate Ksh to
US 2,862,354 859,971 3.3284 0.0050 1,017,900 4,706,807
CG INDEX 6,828,427 17,634,398 0.3872 0.7044 (30,993,595) 44,650,448
IS INDEX (18,456,282) 17,845,894 (1.0342) 0.3186 (56,731,917) 19,819,352
4.3.4.1. The joint effect of CG, IS, & macroeconomic factors on NSSF funding level
A multiple regression was carried out to investigate whether the joint effect of CG index, IS index, &
macroeconomic factors on performance of the NSSF is greater than the individual effect of CG. The results of
the regression indicated that the model explained 94.34% of the variance and that the model was a significant
predictor of pension performance, F(7,14) = 33, p < .001. Whereas the NSE 20 share index (t = 2.24, p<.05) and
Exchange rate (t = 3.33, p<.05) contributed significantly to the model, the other factors did not as indicated by
their low t values and high p values: GDP growth rate (t = 1.36, p=.195); Inflation rate (t = 1.83, p=.089);
Average interest rate (t = -0.318, p=.755); CG index (t = 0.387, p=.704); IS index (t = -1.034, p=.319). The final
predictive model for the NSSF fund was:
NSSF fund value = -185,990,325 + 427,424,473* GDP growth rate + 338,823,452* Inflation rate-
57,278,894*Average interest rate+14,990*NSE 20 share index +
2,862,354*Exchange rate +6,828,427*CG index - 18,456,282*IS index.
The study shows that when one compares the joint effect of CG, IS, GDP growth rate, Inflation rate,
Average interest rate, NSE 20 share index, Exchange rate and individual effect of CG index on performance of
the NSSF, the study findings indicate that the joint effect is greater than the individual effect as shown by the
regression statisticsin table 4 and 24. The tables show that the joint effect performed better as the model
explained 94.34% compared to 67.42% of the later on the variation in the independent variable, the NSSF value.
The Pearson’s correlation coefficient r, a measure of the strength of the association between the two variables,
was 0.97 for the joint effect compared to 0.82 for the individual effect of CG index on NSSF value.
The study also established that the Pearson’s correlation coefficient (r) was greater for the joint effect ranging
between 0.94-0.98 as compared to the individual effect of CG index on the other tested pension funds. For the
later it ranged between 0.50-0.82. Both models were however, significant in predicting the dependent variable,
NSSF value as indicated by the high F valuesand low p values: Joint effect- F (7, 14) = 33, p value<.001;
Individual effect of CG index - F (1,20) = 41.38, p value<.001.
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 64 | Page
4.3.4.2. The joint effect of CG index, IS index, & macroeconomic factors on other pension funds
Table 27:Summary output of the effect of CG, IS, & macroeconomic factors on the performance of
Teleposta pension scheme Regression Statistics
Multiple R 0.9573 R Square 0.9164
Adjusted R Square 0.8745
Standard Error 2,411,255 Observations 22
ANOVA
df SS MS F Significance F
Regression 7 891,721,842,692,605 127,388,834,670,372 22 0.0000016
Residual 14 81,398,097,688,681 5,814,149,834,906
Total 21 973,119,940,381,286 Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (34,837,741) 12,277,715 (2.83748) 0.01317 (61,170,821) (8,504,660)
GDP growth rate 71,724,070 35,640,099 2.01245 0.06382 (4,716,340) 148,164,479
Inflation rate 22,433,624 20,999,924 1.06827 0.30347 (22,606,733) 67,473,982 Average interest rate 47,150,306 20,448,612 2.30579 0.03694 3,292,396 91,008,217
NSE 20 share index 969 761 1.27408 0.22338 (662) 2,600
Exchange rate Ksh to US 367,867 97,549 3.77108 0.00207 158,644 577,090 CG INDEX 4,192,013 2,000,331 2.09566 0.05477 (98,270) 8,482,297
IS INDEX (4,812,462) 2,024,322 (2.37732) 0.03224 (9,154,200) (470,723)
Table 28:Summary output of the effect of CG, IS, & macroeconomic factors on the performance of
LAPTRUST Regression Statistics
Multiple R 0.9693 R Square 0.9396
Adjusted R Square 0.9094
Standard Error 3,185 Observations 22
ANOVA df SS MS F Significance F
Regression 7 2,210,299,246 315,757,035 31 0.00000017
Residual 14 142,007,799 10,143,414
Total 21 2,352,307,045 Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (57,832) 16,217 (3.5662) 0.0031 (92,614) (23,051)
GDP growth rate 26,831 47,075 0.5700 0.5777 (74,134) 127,797 Inflation rate 51,875 27,737 1.8702 0.0825 (7,616) 111,366
Average interest rate 63,719 27,009 2.3592 0.0334 5,790 121,648
NSE 20 share index 2.4148 1.0046 2.4038 0.0306 0.2602 4.5694 Exchange rate Ksh to US 631 129 4.8989 0.0002 355 908
CG INDEX 4,539 2,642 1.7181 0.1078 (1,127) 10,206
IS INDEX (5,634) 2,674 (2.1072) 0.0536 (11,369) 101
Table 29:Summary output of the effect of CG, IS, & macroeconomic factors on the performance of KPLC
Umbrella Regression Statistics
Multiple R 0.9630
R Square 0.9274
Adjusted R Square 0.8911 Standard Error 2,684
Observations 22
ANOVA
df SS MS F Significance F
Regression 7 1,287,710,580 183,958,654 26 0.00000061
Residual 14 100,847,098 7,203,364
Total 21 1,388,557,678 Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (41,163) 13,666 (3.01205) 0.00933 (70,473) (11,852)
GDP growth rate 22,670 39,670 0.57146 0.57674 (62,414) 107,754 Inflation rate 37,119 23,374 1.58801 0.13461 (13,014) 87,252
Average interest rate 43,081 22,761 1.89276 0.07925 (5,736) 91,898
NSE 20 share index 2 1 2.29259 0.03788 0 4 Exchange rate Ksh to US 453 109 4.17630 0.00093 221 686
CG INDEX 5,103 2,227 2.29214 0.03791 328 9,879
IS INDEX (5,918) 2,253 (2.62637) 0.01993 (10,750) (1,085)
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 65 | Page
Table 30:Summary output of the effect of CG, IS, & macroeconomic factors on the performance of KPLC
Individual pension scheme Regression Statistics
Multiple R 0.9687
R Square 0.9384 Adjusted R Square 0.9076
Standard Error 1,473
Observations 22 ANOVA
df SS MS F Significance F
Regression 7 462,995,003 66,142,143 30 0.00000020 Residual 14 30,379,266 2,169,948
Total 21 493,374,270
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (19,298) 7,501 (2.5729) 0.0221 (35,385) (3,211)
GDP growth rate (4,464) 21,773 (0.2050) 0.8405 (51,163) 42,235
Inflation rate 14,488 12,829 1.1293 0.2778 (13,028) 42,004 Average interest rate 24,646 12,492 1.9729 0.0686 (2,147) 51,440
NSE 20 share index 1 0 1.4867 0.1593 (0) 2
Exchange rate Ksh to US 250 60 4.1915 0.0009 122 378 CG INDEX 659 1,222 0.5396 0.5980 (1,962) 3,280
IS INDEX (1,551) 1,237 (1.2543) 0.2303 (4,204) 1,101
Table 31:Summary output of the effect of CG, IS, & macroeconomic factors on the performance of
Heritage GRP pension scheme Regression Statistics
Multiple R 0.9724 R Square 0.9455
Adjusted R Square 0.9182
Standard Error 362 Observations 22
ANOVA df SS MS F Significance F
Regression 7 31,862,922 4,551,846 35 0.00000009 Residual 14 1,837,376 131,241
Total 21 33,700,297 Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (4,063) 1,845 (2.2023) 0.0449 (8,019) (106)
GDP growth rate 9,850 5,355 1.8396 0.0871 (1,634) 21,335
Inflation rate 6,957 3,155 2.2049 0.0447 190 13,724 Average interest rate (4,007) 3,072 (1.3043) 0.2132 (10,597) 2,582
NSE 20 share index 0 0 3.3144 0.0051 0 1
Exchange rate Ksh to US 55 15 3.7749 0.0020 24 87 CG INDEX (392) 301 (1.3050) 0.2129 (1,037) 252
IS INDEX 300 304 0.9852 0.3413 (353) 952
Table 32:Summary output of the effect of CG, IS, & macroeconomic factors on the performance of
Heritage Life Insurance Fund Regression Statistics
Multiple R 0.9460 R Square 0.8949
Adjusted R Square 0.8423
Standard Error 396
Observations 22
ANOVA df SS MS F Significance F
Regression 7 18,725,895 2,675,128 17 0.0000075
Residual 14 2,200,057 157,147
Total 21 20,925,952 Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (2,730) 2,018 (1.3527) 0.1976 (7,060) 1,599
GDP growth rate 6,560 5,859 1.1196 0.2817 (6,007) 19,127 Inflation rate 5,022 3,452 1.4548 0.1678 (2,382) 12,427
Average interest rate (4,632) 3,362 (1.3777) 0.1899 (11,842) 2,579
NSE 20 share index 0.2641 0.1250 2.1122 0.0531 (0.0041) 0.5323 Exchange rate Ksh to US 42 16 2.6195 0.0202 8 76
CG INDEX (571) 329 (1.7369) 0.1044 (1,277) 134
IS INDEX 500 333 1.5030 0.1550 (214) 1,214
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 66 | Page
Table 33:Summary output of the effect of CG, IS, & macroeconomic factors on the performance of
Liberty Life Insurance Fund Regression Statistics
Multiple R 0.971
R Square 0.942 Adjusted R Square 0.914
Standard Error 896
Observations 22 ANOVA
df SS MS F Significance F
Regression 7 183,958,258 26,279,751 33 0.00000013 Residual 14 11,250,265 803,590
Total 21 195,208,523
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (16,468) 4,564 (3.6078) 0.0029 (26,258) (6,678)
GDP growth rate 31,472 13,250 2.3752 0.0324 3,054 59,890
Inflation rate 6,960 7,807 0.8914 0.3878 (9,785) 23,704 Average interest rate 26,024 7,602 3.4232 0.0041 9,719 42,329
NSE 20 share index 0.4733 0.2828 1.6740 0.1163 (0.1331) 1.0798
Exchange rate Ksh to US 169 36 4.6563 0.0004 91 247 CG INDEX 682 744 0.9166 0.3749 (913) 2,277
IS INDEX (1,047) 753 (1.3912) 0.1859 (2,661) 567
Table 34:Summary output of the effect of CG, IS, & macroeconomic factors on the performance of
Britam Life Insurance Fund Regression Statistics
Multiple R 0.9671 R Square 0.9354
Adjusted R Square 0.9031
Standard Error 11,054 Observations 22
ANOVA df SS MS F Significance F
Regression 7 24,759,933,020 3,537,133,289 28.95 0.000000275 Residual 14 1,710,610,207 122,186,443
Total 21 26,470,543,227 Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (123,880) 56,284 (2.2010) 0.0450 (244,597) (3,162)
GDP growth rate 65,486 163,383 0.4008 0.6946 (284,936) 415,909
Inflation rate 61,143 96,269 0.6351 0.5356 (145,333) 267,620 Average interest rate 168,056 93,742 1.7928 0.0946 (32,999) 369,112
NSE 20 share index 5 3 1.4876 0.1590 (2) 13
Exchange rate Ksh to US 1,646 447 3.6818 0.0025 687 2,606 CG INDEX 5,214 9,170 0.5686 0.5786 (14,454) 24,882
IS INDEX (12,085) 9,280 (1.3023) 0.2138 (31,989) 7,819
Table 35: Summary output of the effect of CG, IS, & macroeconomic factors on the performance of CfC
holdings Regression Statistics
Multiple R 0.9597 R Square 0.9211
Adjusted R Square 0.8816
Standard Error 32,412
Observations 22
ANOVA
df SS MS F Significance
F
Regression 7
171,653,909,688 24,521,987,098
23.34 0.00000108
Residual 14
14,707,737,976 1,050,552,713
Total 21
186,361,647,664
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (336,738)
165,038
(2.0404)
0.0606 (690,709) 17,232
GDP growth rate 547,362
479,076 1.1425
0.2724 (480,155) 1,574,878 Inflation rate 470,996 1.6685 (134,440) 1,076,431
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 67 | Page
282,282 0.1174
Average interest rate 151,548 274,871 0.5513
0.5901 (437,993) 741,088
NSE 20 share index 15
10 1.4848
0.1598 (7) 37 Exchange rate Ksh to
US 4,411
1,311 3.3637
0.0046 1,598 7,223
CG INDEX 56,604 26,889 2.1051
0.0538 (1,066) 114,275
IS INDEX
(67,728)
27,211
(2.4890)
0.0260 (126,090)
(9,366)
Table 36: Summary output of the effect of CG, IS, & macroeconomic factors on the performance of CfC
Life assurance Regression Statistics
Multiple R 0.9646
R Square 0.9305
Adjusted R Square 0.8958
Standard Error 1,240
Observations 22
ANOVA
df SS MS F Significance F
Regression 7 288,253,466 41,179,067 26.78 0.000000453
Residual 14 21,529,470 1,537,819 Total 21 309,782,936
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (16,459) 6,314 (2.6066) 0.0207 (30,002) (2,916)
GDP growth rate 22,933 18,329 1.2512 0.2314 (16,379) 62,246 Inflation rate 13,536 10,800 1.2533 0.2306 (9,628) 36,700
Average interest rate 23,653 10,517 2.2491 0.0411 1,097 46,209
NSE 20 share index 0.2825 0.3912 0.7221 0.4821 (0.5565) 1.1214 Exchange rate Ksh to US 198 50 3.9496 0.0015 91 306
CG INDEX 309 1,029 0.3005 0.7682 (1,897) 2,516
IS INDEX (947) 1,041 (0.9095) 0.3785 (3,180) 1,286
Table 37:Summary output of the effect of CG, IS, & macroeconomic factors on the performance of CIC
Life assurance Regression Statistics
Multiple R 0.9787
R Square 0.9579 Adjusted R Square 0.9368
Standard Error 2,884
Observations 22 ANOVA
df SS MS F Significance F
Regression 7 2,645,996,918 377,999,560 45.46 0.000000015
Residual 14 116,414,185 8,315,299 Total 21 2,762,411,103
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (49,677) 14,683 (3.3833) 0.0045 (81,169) (18,185) GDP growth rate 43,403 42,622 1.0183 0.3258 (48,012) 134,819
Inflation rate 37,950 25,114 1.5111 0.1530 (15,914) 91,813
Average interest rate 70,761 24,455 2.8936 0.0118 18,312 123,211
NSE 20 share index 1.5928 0.9096 1.7512 0.1018 (0.3580) 3.5437
Exchange rate Ksh to US 592 117 5.0730 0.0002 342 842
CG INDEX (110) 2,392 (0.0460) 0.9640 (5,241) 5,021 IS INDEX (1,951) 2,421 (0.8059) 0.4338 (7,143) 3,241
Table 38:Summary output of the effect of CG, IS, & macroeconomic factors on the performance of
Jubilee Life assurance Regression Statistics
Multiple R 0.9737
R Square 0.9482 Adjusted R Square 0.9223
Standard Error 10,017
Observations 22 ANOVA
df SS MS F
Significance
F
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 68 | Page
Regression 7
25,697,927,663 3,671,132,523
36.59 0.000000061
Residual 14
1,404,746,463 100,339,033
Total 21
27,102,674,125
Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept (134,851)
51,005
(2.6439)
0.0193
(244,245)
(25,457)
GDP growth rate 108,911
148,058 0.7356
0.4741
(208,641) 426,463
Inflation rate 85,285 87,239 0.9776
0.3449
(101,824) 272,394
Average interest rate 169,985 84,948 2.0010
0.0652
(12,211) 352,182
NSE 20 share index 5.1431
3.1596 1.6278
0.1259
(1.6335) 11.9198 Exchange rate Ksh to
US 1,801
405 4.4440
0.0006 932 2,670
CG INDEX 3,231 8,310 0.3888
0.7033
(14,592) 21,053
IS INDEX
3333(9,532)
8,410
(1.1335)
0.2761
(27,568) 8,505
Tables 25-36 show the study findings of the joint effect of CG index, IS index and macroeconomic
factors on performance of other evaluated pension fundsnamely LAPTRUST, KPLC Umbrella, KPLC
Individual, Heritage GRP, Heritage Life Insurance Fund, Liberty Life Insurance Fund, Britam Life Insurance
Fund, CfC holdings, CfC Life assurance, CIC Life assurance, Jubilee Life assurance and Jubilee Life
assurance.The estimated R2 for the joint effect ranged between 89-97%, compared to the individual effect of CG
on respective pension fund values, which varied between 25-67% as indicated by tables 4-9. The joint effect
thus explains 89-96% of the variation in the dependent variable, pension funding level compared to the
individual effect of CG index that only explained only 25-67%. The joint effect was therefore greater.Moreover,
the F values for the joint effect were higher and ranged between 17-47 with p values < .001, indicating that the
regression model was statistically significant in predicting the funding levels. In comparison, the individual
effect of CG index on pension funding had lower F values ranging between -6.43-37 with p values <.001 with
exception of Teleposta pension scheme, F (1,20) = 6.78, p<.05. The findings suggest that the joint effect had
more significant models with p values<.001 than the individual effect of CG on pension performance. The
models therefore had explanatory power. This implies that the joint effect of CG, IS, & macroeconomic factors
on pension fund value is greater than the individual effect of CG.
The study established that the T statistic of the joint effect showed several independent variables were
statistically significant in the prediction of fund levels. Their influence however varied with the fund being
involved as indicated below:
Table 39: T Statistics Scheme No. ofSignificant
independent
variable
Significant
independent variable
Level of Significance
1. NSSF 2 NSE 20 share index T statistic: t = 2.23, p value <.05
Exchange rate T statistic: t = 3.33, P value < .05
2. Teleposta pension scheme 4 Average Interest Rate T statistic: t = 2.3, p value < .05 Exchange rate. T statistic: t = 3.8, p value < .05
CG index. T statistic: t = 2.1, p value < .05
IS index T statistic: t = -2.4, p value < .05 3. LAPTRUST 4 Average Interest Rate T statistic: t = 2.4, p value < .05
NSE 20 share index T statistic: t = 2.4, P value < .05
Exchange rate T statistic: t = 4.9, p value < .001 IS index T statistic: t = -2.1, P value < .05
4. Kenya Power Company
DC
4 NSE 20 share index T statistic: t = 2.3, p value < .05
Exchange rate T statistic: t = 4.2, P value < .001
CG index T statistic: t = 2. 3, p value < .05
IS index. T statistic: t = -2.6, P value < .05
5. KPLC (Individual.) 1 Exchange rate T statistic: t = 4.2, P value < .001. 6. Heritage (Grp) 3 Inflation rate. T statistic: t = 2.2, p value < .05
NSE 20 share index. T statistic: t = 3.3, p value < .05
Exchange rate. T statistic: t = 3.8, p value < .05 7. Heritage Life insurance
Fund
2 NSE 20 share index. T statistic: t = 2.1, p value < .05
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 69 | Page
Scheme No. ofSignificant
independent
variable
Significant
independent variable
Level of Significance
Exchange rate T statistic: t = 2.6, p value < .05
8. Liberty Life insurance
Fund
3 GDP growth rate T statistic: t = 2.4, p value < .05
Average Interest Rate T statistic: t = 3.4, p value < .05
Exchange rate T statistic: t = 4.7, p value < .001
9. Britam Life assurance 1 Exchange rate T statistic: t = 3.7, p value < .05 10. CFC Holdings 3 Exchange rate T statistic: t = 3.4, p value < .05
CG index T statistic: t = 2.1, p value < .05
IS index T statistic: t = -2.5, p value < .05 11. CFC life assurance 2 Average Interest Rate. T statistic: t = 2.2, p value < .05
Exchange rate T statistic: t = 3.9, p value < .001
12. Jubilee Life Insurance fund
1 Exchange rate. T statistic: t = 4.4, p value < .001
13. CIC life assurance fund 2 Average Interest Rate T statistic: t = 2.9, p value < .05
Exchange rate T statistic: t = 5.1, p value < .001
The study findings suggest that the joint effect of CG index, IS index & macroeconomic variables on pension
performance is greater than the individual effect of CG on pension performance in Kenya.
4.4. Conclusion and recommendations
The pension industry in Kenya has undergone tremendous growth since independence to the level that
it now manages over Sh1 trillion worth Assets. Despite the operationalisation of the Retirement Benefits Act
(RBA) of 1997 and its subsequent amendments to regulate, supervise and promote the retirement benefits
schemes, and therefore prevent account holders against losses, the funds are nonetheless still exposed to
investment risks that are related with economy as a whole such as market risk/systemic risks and specific
risk/unsystematic risks (specific to a particular company or industry).The Markowitz’s Modern Portfolio Theory
(MPT) is one of the tools used by pension fund managers to manage investment risks. The MPT, a theory of
finance on how risk-averse investors attempt to maximize diversification in investing, with the aim of selecting
a collection of assets that optimize or maximize expected return based on a given level of market risk,
emphasizing that risk is an inherent part of higher reward. This portfolio combination allows investors to utilize
risky and un-risky assets.
The study investigated the effect of macroeconomic factors, (GDP growth rate, inflation rate, interest
rate, NSE 20 share index, exchange rate), corporate governance, investment strategy on the level of funding of
pension funds during the study period of 1997 to 2018. The study findings show that there is a significant
relationship between funding level and these factors as shown by the high Pearson’s correlation coefficient of
>.7, the high R squared and the adjusted R squares values and the low p values. Tables 26-38 show that the
entire regression modelson joint effects for the various pension funds were statistically significant in predicting
the funding levels because the p values <.001. This suggests that the models had explanatory powern.
The regression coefficients which,isolates the role of one variable from all of the others in the model,
established that the level of influence of each individual independent factor on the dependent variable,pension
funding level, varied with the scheme being evaluated. The exchange rate was individually significant in the
prediction of the the NSSF and Jubilee fund values. Average interest rate and IS index were individually useful
in the prediction of the Teleposta fund while exchange rate and average interest rate were significant for the
LAPTRUST and the Kenya Power Pension fund (DB). For the case of Kenya Power Pension fund (DC),
exchange rate, average interest rate, CG and IS indices were individually significant in the prediction of the fund
value. This implies that these factors have a relationship with the financial performance of pension funds and
hence the accumulation of retirement benefits.
The study findings suggest that different risk factors in the investment markets, whether systemic or
un-systemic need to be taken into consideration when making investment management decisions. The study
showsthat there is a variation in the predictor’s individual usefulness in the prediction of the fund value. The
results imply that knowledge of both systematic and unsystematic risk factors is critical in the management of
investments of various pension schemes. Investment management should therefore take into consideration the
different risk factors when making investment decisionsguided by the Modern Portfolio Theory (MPT).
The results of this study are significant as they providecritical information concerning key investment
risk factors,whether systemic and unsystematic in Kenya to the players in the pension industry, particularly
existing or potential members, pension managers, policy makersand the government, to make investment
decisions that will determine pension performance. Knowledge of the different types of investment riskswill be
critical in mitigation and minimization of risk with portfolio management tools. The investment process is vital
in contributing to the generation of adequate pension funds.
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 70 | Page
The study nonetheless faced a number of challenges during its undertaking. Key among them included
lack of data from the pension funds particularly in the early years of study; lack of cooperation and time by
management during the administration of the questionnaire; reforms undertaken by some of the funds during the
study period. Future issues to consider in similar studies is to utilize monthly or quarterly data returns for both
the macroeconomic and funding values rather than the annual figure. To ensure exhaustive study, one could also
concentrate of a limited number of pension funds.
References [1]. Aguinis, H., & Glaves, A. (2012), What we know and dont know about corporate social responsibility: A review and research.
Journal of Management March Vol. 38 N° 4, July 2012
[2]. Aguilera, R. V., I. Filatotchev, et al. (2008). An organizational approach to comparative corporate governance: Costs, contingencies and complementarities. Organization Science 19(3): 475-492.
[3]. Ambachtsheer, K. P., Capelle, R., Lum, H. (2006). Pension fund governance today: Strengths, weaknesses, and opportunities for
improvement, Financial Analysts Journal. [4]. Antolín, P., Stewart F. (2009). Private pensions and policy responses to the financial and economic crisis, OECD Working Papers
on Insurance and Private Pensions No. 36. Organisation for Economic Co-operation and Development, Paris.
[5]. Asher, M. & Nandy, A. (2006). Reforming provident and pension fund regulation in India. Journal of Financial Regulation and
Compliance.
[6]. Baron, R.M., & Kenny, D.A. (1986). The moderator-mediator variable distinction in social psychological research: conceptual,
strategic, and statistical considerations, J Pers Soc Psychol. 51(6): 1173-82. [7]. Bebchuk, L.A & Weisbach, M.S. (2010). The state of corporate governance research. Review of Financial Studies. 23 (3) 939-961.
[8]. Besley, T. & Prat, A. (2005). Pension fund governance and the choice between Defined Benefit and Defined Contribution plans. CEPR Discussion Paper No. 3955.
[9]. Bhagat, S., & B. Black. (2002). The non-correlation between board independence and long-term firm performance, Journal of
Corporation Law 27,23 1-274. [10]. Black, B. S., H. Jangand W. Kim (2002). Does corporate governance matter? Evidence from the Korean Market', Working Paper,
Stanford Law School, Korea University and KDI School of Public Policy and Management.
[11]. Blake, D., B. Lehmann & A. Timmermann. (1999). Asset allocation dynamics and pension fund performance. Journal of Business 72, 429-461.
[12]. Bodie, Z. (1990b). Pension funds and financial innovation, Financial Management, 11-21.
[13]. Brown, G. P. Draper & E. McKenzie (1997). Consistency of UK pension fund performance. Journal of Business Finance and Accounting 24, March, 155-178.
[14]. Chan, K.C., Karceski, J., & Lakonishok, J. (1998). The risk and return from factors, Journal of Financial and Quantitative Analysis,
33, 159-188. [15]. Davis, E. P., & Hu, Y. (2008). Pension Fund Investment and Regulation. A Macro Study. Working Paper No. 06-11, Economics
and Finance Section, Brunel University.
[16]. Donaldson, L. & Davies, J. H. (1989, 1991). Stewardship Theory or Agency Theory. CEO governance and shareholder returns. Australian Journal of Management 16: 49-64.
[17]. Eisenhardt, K. M. (1989). Agency Theory: An Assessment and Review. Academy of Management Review 14(1): 57–74.
[18]. Epps, R. W., & Cereola, S. J. (2008). Do institutional shareholder services (ISS) corporate governance ratings reflect company's operating performance? Critical Perspectives on Accounting, 19(8), 1135-1148.
[19]. Fama, E. F., & Jensen M., C. (1983a). Agency problems and residual claims. Journal of Law and Economics, 26(2): 327–49.
[20]. Farrar, Donald E.; Glauber, Robert R. (1967). Multicollinearity in regression analysis: The problem revisited. Review of Economics and Statistics. 49 (1): 92–107. doi:10.2307/1937887. JSTOR 1937887.
[21]. Fich, E. M., & Shivdasani, A., (2006). Are busy boards effective monitors? The Journal of Finance, 61(2), 689–724.
[22]. Flannery, M.J. & Protopapadakis, A.A. (2002). Macroeconomic factors do influence aggregate stock returns, Review of Financial Studies, 15, 751-782.
[23]. Fosberg, R. H., (1989). Outside directors and managerial monitoring, Akron Business and Economic Review 20,24-32.
[24]. Freeman, E.: 1984. Strategic Management: A Stakeholder Approach (Pitman, Boston). [25]. Government of Kenya, (1997). The Retirement Benefits Act, No. 3 of 1997.
[26]. Hinz, R., Rudolph, H., Antolin, P., & Yermo, J. (2010). Evaluating the financial performance of pension funds. The World Bank,
Washington. [27]. Heracleous, L. (2001). What is the impact of corporate governance on organizational performance? Corporate Governance: An
International Review 9,165-173.
[28]. Hillman, A.J. and Keim, G.D. (2001). Shareholder value, stakeholder management, and social issues: What’s the bottom line?
Strategic Management Journal, 22, 125-139.
[29]. Humpe, A., & Macmillan, P. (2007). Can macroeconomic variables explain long term stock market movements? A Comparison of
the US and Japan, University of St Andrews. [30]. Jensen, M.C., &Meckling, W. (1976). Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure. Journal of
Financial Economics. 3, 305-360
[31]. Malik, U., Khan, A., & Tanveer, S. (2016) An empirical analysis of corporate governance and firm value: Evidence from KSE-100 Index,Department of Management Sciences, University of Haripur, Pakistan.
[32]. Kwon, C.S., & Shin, T.S. (1999). Co-integration and causality between macroeconomic variables and stock market returns. Global
Finance Journal, 10 (1), 71-81. [33]. Lishenga, L. (2011). Corporate governance reaction to declining firm performance: Evidence from the NSE. International Journal
of Governance, 1, 1.
[34]. Maher, M. & Andersson T., (2000). Corporate Governance: Effects on Firm Performance and Economic Growth, Convergence and Diversity of Corporate Governance Regimes and Capital Markets, Oxford University Press, 2000.
[35]. Manuel, A., & Andreas, Z. (2008). Performance and governance of Swiss pension funds University of St. Gallen, Swiss Institute of
Banking and Finance, Rosenberg Strasse 52, 9000 St. Gallen, Switzerland. [36]. Markowitz, H. (1952). Portfolio selection. The Journal of Finance, 7(1), 77-91.
Effect of Corporate Governance, Investment Strategyand Macroeconomic Factors on Financial ..
DOI: 10.9790/487X-2204083171 www.iosrjournals.org 71 | Page
[37]. Mutegi, I. (2014). Effects of corporate governance activities on organizational performance of selected occupational retirement
schemes in Kenya, vol. 2 (13), pp 237-255, The Strategic Journal of Business and Change Management. [38]. Njuguna, A.G. (2011). Determinants of Pension Governance: A Survey of Pension Plans in Kenya. International Journal of
Business and Management, 6(11).
[39]. Ochieng, D. E., & Oriwo, E. A. (2012). The relationship between macroeconomic variables and stock market performance in Kenya. DBA Africa Economic Review, Vol 3 (1), 38-49.
[40]. OECD (2009). OECD guidelines on pension fund asset management. OECD Working Papers on Insurance and Private Pensions 8,
OECD Publishing, Paris, France. [41]. Olweny, T. & Omondi, K. (2011). The effect of macroeconomic factors on stock return volatility in the Nairobi stock exchange,
Kenya. Economics and Finance Review, 1(10), 34 – 48. Available online at http://www.businessjournalz.org/efr
[42]. Ongore, V. O., & Kobonyo, P. (2011). Effects of selected corporate governance characteristics on firm performance. Empirical evidence from Kenya. International Journal of Economics and Financial Issues, Eco Journals, 1(3), 99-122.
[43]. RBA (2018). Retirement Benefits Industry Performance Report for the year 2017.
[44]. Ross, S. A. (1976). The arbitrage theory of capital asset pricing, Journal of Economic Theory, 13, 341-360. [45]. Sharpe, W.F. (1964). Capital Asset Prices: A theory of market equilibrium under conditions of risk. Journal of Finance, vol. 18, pp.
425-442.
[46]. Shleifer, A. and R. Vishny (1997). A survey of corporate governance, Journal of Finance, 52(2), pp. 737-783. [47]. Stewart F. (2010). Pension funds’ risk-management framework: regulation and supervisory oversight. OECD working paper on
insurance and private pensions.
[48]. Tapia, & Waldo (2008a). Comparing aggregate investment returns in privatively managed pension funds. OECD working paper on insurance and private pensions, no. 22, Paris.
[49]. Tonks, I., (2006). Pension fund management and investment performance, (23), pp. 456-480. Oxford Handbook of Pensions and
Retirement Income, Oxford University Press. [50]. Useem, M., & Mitchell, O. S., (2008). Holders of the purse strings: Governance and Performance of Public Retirement Systems.
Social Science Quarterly, 81(2). 489 – 506.
[51]. Vittas, D., Impavido G., & O’Connor, R. (2008). Upgrading the Investment Policy Framework of Public Pension Funds’, Policy Research Working Paper No. 4499. Washington D. C.: World Bank.
[52]. Walker, E. & Iglesias A. (2010). Financial performance of pension fund systems around the world: An exploratory study. World
Bank, p. 39-94. [53]. Wichers, C. R. (1975). The Detection of Multicollinearity: A Comment. Review of Economics and Statistics. 57 (3): 366–368.
AkwimbiAmbaka William. “Effect of Corporate Governance, Investment Strategyand
Macroeconomic Factors on Financial Performance of Pension Schemes in Kenya."
IOSR Journal of Business and Management (IOSR-JBM), 22(4), 2020, pp. 31-71.